The Fast atmOspheric traCe gAs retrievaL (FOCAL) algorithm has been originally developed to derive XCO2 from OCO-2 measurements (Reuter et al., 2019a.b). Later, the FOCAL method has also been successfully applied to measurements of the Greenhouse gases Observing SATellites GOSAT and GOSAT-2. (Noël et al., 2021).

FOCAL has proven to be a fast and accurate retrieval method well suited for the challenges of forthcoming greenhouse gas missions producing large amounts of data. FOCAL is one of the foreseen operational algorithms for the forthcoming CO2M mission. The FOCAL retrieval results delivered by the University of Bremen are the baseline for the new GOSAT XCO2 products of the Copernicus Atmosphere Monitoring Service (CAMS).

In this presentation we will show recent results from GOSAT and GOSAT-2 FOCAL retrievals for XCO2 and other gases, e.g. methane (XCH4, both FP products and proxy), water vapour (XH2O) and HDO (δD). For GOSAT-2, we will also present results for carbon monoxide (XCO) and nitrous oxide (XN2O). This will include comparisons with independent data sets.

References:

Noël, S., M. Reuter, M. Buchwitz, J. Borchardt, M. Hilker, H. Bovensmann, J. P. Burrows, A. Di Noia, H. Suto, Y. Yoshida, M. Buschmann, N. M. Deutscher, D. G. Feist, D. W. T. Griffith, F. Hase, R. Kivi, I. Morino, J. Notholt, H. Ohyama, C. Petri, J. R. Podolske, D. F. Pollard, M. K. Sha, K. Shiomi, R. Sussmann, Y. Té, V. A. Velazco and T. Warneke, XCO2 retrieval for GOSAT and GOSAT-2 based on the FOCAL algorithm, Atmos. Meas. Tech., 14(5), 3837-3869, 2021, doi: rm10.5194/amt-14-3837-2021. URL https://amt.copernicus.org/articles/14/3837/2021/

Reuter, M., M. Buchwitz, O. Schneising, S. Noël, V. Rozanov, H. Bovensmann and J. P. Burrows, A fast atmospheric trace gas retrieval for hyperspectral instruments approximating multiple scattering - part 1: Radiative transfer and a potential OCO-2 XCO2 retrieval setup, Rem. Sens., 9(11), 1159, 2017a, ISSN 2072-4292, doi: rm10.3390/rs9111159. URL http://www.mdpi.com/2072-4292/9/11/1159

Reuter, M., M. Buchwitz, O. Schneising, S. Noël, H. Bovensmann and J. P. Burrows, A fast atmospheric trace gas retrieval for hyperspectral instruments approximating multiple scattering - part 2: Application to XCO2 retrievals from OCO-2, Rem. Sens., 9(11), 1102, 2017b, ISSN 2072-4292, doi: rm10.3390/rs9111102. URL http://www.mdpi.com/2072-4292/9/11/1102

Anthropogenic emissions from cities and power plants contribute significantly to air pollution and climate change. Their emission plumes are visible in satellite images of atmospheric trace gases (e.g. CO₂, CH₄, NO₂, CO and SO₂) and data-driven approaches are increasingly being used for quantifying the sources.

We present an open-source software library written in Python for detecting and quantifying emissions in satellite images. The library provides all processing steps from the pre-processing of the satellite images, the detection of the plumes, the quantification of emissions, to the extrapolation of individual estimates to annual emissions. The plume detection algorithm identifies regions in satellite images that are significantly enhanced above the background and assigns them to a list of potential sources such as cities, power plants or other facilities. Overlapping plumes are automatically detected and segmented. The plume shape is described by a set of polygons and a centerline along the plume ridge. Functions are available for converting geographic coordinates (longitude and latitude) to along- and across-plume coordinates. The emissions can be quantified using various data-driven methods such as computing cross-sectional fluxes or fitting a Gaussian plume model. The models can account for the decay of, for example, NO₂ downstream of the source. Furthermore, it is possible to fit two species simultaneously (e.g. CO₂ and NO₂) to constrain the shape of the CO₂ plume using NO₂ observation that typically have better accuracy. Annual emissions can be obtained by fitting a periodic C-spline to a time series of individual estimates.

A tutorial is available using Jupyter Notebooks to introduce the features of the library. Examples are demonstrated for Sentinel-5P NO₂ observations and for synthetic CO₂ and NO₂ satellite observations available for the CO2M satellite constellation. The library and its tutorial are available on Gitlab (https://gitlab.com/empa503/remote-sensing/ddeq) and can conveniently be installed using Python's package installer:

python -m pip install ddeq

The library is licensed under the "GNU Lesser General Public License" and can therefore be used in both open-source and proprietary software. Interested users are encouraged to contribute to the development of the library by reporting bugs, requesting or implementing new features and applying the library for detecting and quantifying emission plumes. If you are interested in contributing to the development of the software library, please contact the developers.

Methane is one of the most powerful greenhouse gases (GHG), having 84 more times warming potential than carbon dioxide. According to the last IPCC AR6 report, a strong, rapid and sustained reduction of GHG emissions would limit the warming effect and improve air quality. The 20% of the global methane emissions come from the fossil fuel industry. These have a direct implication in the global warming equivalent to 0.1ºC out of the 0.5º C globally attributed to methane.
TROPOMI, the TROpospheric Monitoring Instrument on board of the Sentinel-5P, can play a key role in tackling methane emissions in the largest oil and gas producing region of the United States, the Permian basin.
During the COVID-19 lockdown in 2020, TROPOMI was able to capture the reduction of maximum values of methane tropospheric concentrations in the two most productive sub-basins (Delaware and Midland) and the increase of the minimum and average values in both. With the latest changes in the algorithm, the methane retrievals from TROPOMI have improved, not only spatially but also temporally, increasing the spatial coverage of the Permian basin on a daily basis.
In order to illustrate the implications of the new algorithm in the use of TROPOMI for methane emissions, different cases showing plumes in the Permian basin have been studied with the support of Sentinel-2. Using the ratio between bands 12 and 11 on the day of the detected plume and the median of the scene from one month before and one month after, it has been possible to compare the new methane retrievals obtained with TROPOMI versus the retrievals obtained with Sentinel-2. The use of Sentinel-2 in the Permian basin, which is a difficult area in terms of source identification due to the crowdedness of O&G facilities, has also returned a diverse list of false methane retrievals obtained with the band ratios of Sentinel-2.
The comparation of the different possible sources detected with Sentinel-2 and the retrievals of TROPOMI, show a spatial relationship with the sources identified over flare stacks, which do not light on the day of the detected plume but usually doing it. Other sources identified, e.g., flare stacks lighting on the day of the identified plume were also located on the same area or in the surroundings, but in less quantity.
The improvement of the methane algorithm on TROPOMI will play a crucial role in the development of cost efficient LDAR (Leak Detection And Repair) activities, reducing the area of source location, increasing the time response and reducing the methane release to the atmosphere.

Using satellite data for estimating carbon dioxide (CO2) emissions from anthropogenic sources has become increasingly important since the Paris Agreement was adopted in 2015, due to their global coverage. The very first study that estimated CO2 emissions from individual power plants using satellite data was published in 2017 (Nassar et al., 2017). In recent years, the literature has been rapidly expanding with several new approaches and case studies. Many of the proposed techniques for estimating CO2 emissions from local sources are based on single satellite overpasses (e.g., Varon et al., 2018). To estimate nitrogen oxide (NOx) emissions from averaged NO2 columns, statistical methods (i.e., based on multiple spatially co-located observations) are often applied. In Europe, one of the key activities to respond to Paris Agreement’s goal to monitor anthropogenic CO2, is the Copernicus Carbon Dioxide Monitoring mission (CO2M).

In this work, we discuss the use of statistical methods for estimating the CO2 emissions. The advantage of the statistical methods is that they do not require complex atmospheric modeling, and they generally provide more robust emission estimates compared to individual satellite overpasses. In addition, these methods have been successfully applied to instruments and locations, where the individual plumes are not detectable, but the emission signal becomes visible when multiple scenes are averaged. In particular, we use divergence method, developed originally for NO2 by Beirle et al. (2019), to estimate CO2 emissions, from the synthetic SMARTCARB dataset (Kuhlmann et al., 2020) that has been created in order to prepare for the upcoming CO2M mission. We analyze the effect of different denoising techniques to the CO2 emission estimates. In addition, we estimate source specific NOx-to-CO2 emission ratio and discuss converting the estimated NOx emissions to CO2 emissions.

Methane is the world's second most important anthropogenic greenhouse gas. It is rising as a result of a variety of factors, including agriculture (e.g., livestock and rice production) and energy generation (mining and use of fuel). Some natural processes, such as the release of methane from natural wetlands, have also changed as a result of human intervention and climate change.
An important uncertainty in the modelling of methane emissions from natural wetlands is the wetland area. It is difficult to model because of several factors, including its spatial heterogeneity on a large range of scales. As we demonstrate using simulations spanning a large range in resolution, getting the spatiotemporal covariance between the variables that drive methane emissions right is critical for accurate emission quantification. This is done using a high-resolution wetland map (100x100m²) and soil carbon map (250x250m²) of the Fenno-Scandinavian Peninsula, in combination with a highly simplified CH₄ emission model that is coarsened in six steps from 0.005° to 1°.
We find a strong relation between wetland emissions and resolution (up to 12 times higher CH₄ emissions for high resolution compared to low resolution), which is sensitive, however, to the sub-grid treatment of the wetland fraction.
As soil moisture is likely to have strong controlling effects on temporal and spatial variability in CH₄ emissions from wetlands. We try to improve CH₄ emissions using high-resolution remote sensing soil moisture datasets, in comparison to modelled soil moisture datasets obtained from the global hydrological model PCR-GLOBWB (PCRG). FluxNet CH₄ observations for 9 selected sites spread over the northern hemisphere were used to validate our simplified model results over the period between 2015 and 2019. As we will show, realistic estimates can be obtained using a highly simplified representation of CH₄ emissions at a high resolution, which is a promising step for minimizing the significant uncertainties for the modelling of CH₄ emissions at local and regional scales.

The global growth rate of methane in the atmosphere shows large fluctuations, the explanation of which has been a major source of controversy in the scientific literature. The renewed methane increase after 2007 has been attributed to either natural or anthropogenic sources, with the latter either dominated by agricultural or fossil emissions. Interannual variability in the hydroxyl radical, the main atmospheric sink of methane, has also been proposed as the dominant driver of the temporary pause in the methane increase prior to 2007. The average of atmospheric methane over the past 5 years is the highest since its atmospheric measurements started in the mid 1980s, with record high growth in 2020 despite the pandemic. As a result, methane is by far the largest contributor to the departure of the path to the 2oC target. Again, the exact causes for this record high growth is up for discussion, where it is clear that also the role of OH needs to be considered.

This shows that atmospheric monitoring of methane is needed, but also that the current capabilities are still insufficient to provide conclusive answers about its global drivers. One of the ways to better address this is to try to better resolve the 3D distribution of methane in the atmosphere, realising that the sinks and sources have a different vertical distribution.
Methane has been measured successfully from space using both SWIR and TIR observations. Recently the TROPOMI instrument has made a huge step in SWIR observations from space, and IASI sensors have been providing TIR observations for over a decade now and will continue to do so. Inverse modeling trying to resolve global sources (and sometimes also sinks) using satellite data measurements have been done, mostly using the SWIR for methane. However, SWIR and TIR have a very different height sensitivity for methane in the atmosphere which in principle – combined - should provide us with a better-resolved 3D distribution of methane and thereby with a better handle on OH as well.
The ESA METHANE+ project aims at using both TROPOMI SWIR and IASI TIR measurements to better disentangle the sources and sink of methane. The project on one hand puts effort in improving the respective satellite data products, while on the other hand focuses on using both datasets in an inverse modeling framework. We will present an overview of the project.

FengYun-3D(FY-3D) is China's polar-orbiting meteorological satellite, launched in November 2017, and has upgraded with enhanced a new payload known as Greenhouse-gases Absorption Spectrometer (GAS) for monitoring CO2, CH4, CO, and N2O. The primary purpose of the GAS/FY-3D is to estimate emissions and absorptions of the greenhouse gases on a sub-continental scale (several thousand kilometers square) more accurately and to assist environmental administration in evaluating the carbon balance of the land ecosystem and making assessments of regional emissions and absorptions. GAS is an instrument that utilizes optical interference to get high spectral resolution of 0.2 cm-1. As a basic character of GAS, the signal to noise ratio (SNR), spectral response, and also the instrumental line shape (ILS) function has been tested on orbit for nearly 8 months in 2018. They all meet the requirements except the SNR in 0.76um band that are affected by micro-vibration effect on orbit.
As the column-averaged mole fraction XCO2 is:
XCO2=0.2095*(CO_2)/(O_2) （1）
CO_2 is the retrieved absolute CO2 column (in molecules/cm2)，O_2 is the retrieved absolute O2 column (in molecules/cm2). 0.2095 is the assumed (column averaged) mole fraction of O2, which used to convert the O2 column into a corresponding dry air column.

As we known, Oxygen is an accurate proxy for the air column because its mole fraction and has negligibly small variations.For removing the influence O2-A band of GAS in retrieving XCO2, the absolute O2 column retrieved by GOAST are used to replace GAS. In order to analyze this uncertainty, the spatiotemporal interpolation results of GOSAT absolute O2 column are compared with OCO-2. Finally, the XCO2 retrieved by GAS are compared with TCCON.

We present our latest results towards the retrieval of methane (CH4) and carbon dioxide (CO2) concentrations on small (local) and large scales using short-wave infrared (SWIR) observations from airborne and space borne sensors.

The code developments are based on Py4CAtS (Python for Computational Atmospheric Spectroscopy), a Python reimplementation of GARLIC, the Generic Atmospheric Radiative Transfer Line-by-line Infrared Code coupled to BIRRA (Beer InfraRed Retrieval Algorithm). BIRRA-GARLIC has recently been validated with TCCON (Total Carbon Column Observing Network) and NDACC (Network for the Detection of Atmospheric Composition Change) ground based measurements.

The software suite BIRRA-Py4CAtS utilizes line data from latest spectroscopic databases such as the SEOM–IAS (Scientific Exploitation of Operational Missions–Improved Atmospheric Spectroscopy) and includes parameterization for Rayleigh and aerosol extinction. Moreover, the latest Py4CAtS version accounts for continuum absorption by means of collision induced absorption (CIA) and facilitates a wide variety of analytical and tabulated instrument spectral response functions. Current developments of the inverse solver BIRRA are directed towards the physical approximation of atmospheric scattering and co-retrieval of effective scattering parameters in order to account for light path modifications when estimating small scale CO2 or CH4 variations.

Methane retrieval results are shown for SWIR observations acquired on a local scale by an airborne HySpex sensor during the CoMet (CO2 and Methane, see Atm. Meas. Tech. special issue) campaign. The retrieval of carbon dioxide is assessed with GOSAT (Greenhouse Gases Observing Satellite) observations. Synthetic/simulated spectra are examined to study the sensitivity of various retrieval setups.

An increasing fleet of space-based, Earth-observing instruments provides coverage of the distribution of the greenhouse gas methane across a range of spatio-temporal scales. In this work, we focus on the synergy between TROPOMI and Sentinel-2 over active oil and gas production regions in Algeria. TROPOMI provides daily global coverage of methane at 7×5.5 km2 resolution, and one of its primary applications is to constrain the global methane distribution. Sentinel-2 provides global coverage every few days at 30m resolution, but with only a few broad spectral bands it is only able to inform on the largest methane point source signals.
In the TROPOMI data over eastern Algeria, large methane plume signals have been detected, most likely coming from large point sources. It is difficult to trace the source location of the plumes based only on TROPOMI, due to its comparatively coarse resolution. Instead, we employ the high-resolution Sentinel-2 data to trace the source locations of these super-emitters to a facility-level. The point source locations are then combined with a generic bottom-up emission inventory, and used as input for a TROPOMI inversion with the Weather Research and Forecasting (WRF) Model to estimate 2020 emissions from Algerian oil and gas fields. Thus, the Sentinel-2 data allows us to track down point source locations, while TROPOMI data provides the best integrated emission quantification of the entire region. In this novel approach, we show that we can optimize both point sources and more diffuse emissions in one systematic framework. In this way, we generate a full emission characterization of the region, in which we estimate the individual contribution of emissions from super-emitters and from diffuse emissions to the methane emission total. This unique information is highly valuable for developing efficient mitigation measures that target oil and gas methane emissions, and by extent their impact on the global climate.

CO2 (carbon dioxide) is the most important anthropogenic greenhouse gas and driving global climate change. Despite this, there are still large uncertainties in our understanding of anthropogenic and natural carbon fluxes to the atmosphere. Satellite observations of the Essential Climate Variable CO2 have the potential to significantly improve this situation. Therefore, a key objective of ESA’s GHG-CCI+ project is to further develop satellite retrieval algorithms needed to generate new high quality satellite-derived XCO2 (column-averaged dry-air mole fraction of atmospheric CO2) data products. One of these algorithms is the fast atmospheric trace gas retrieval FOCAL for OCO-2. FOCAL has been applied also to other satellite instruments (e.g., GOSAT and GOSAT-2) and its development is co-funded by EUMETSAT as it is a candidate algorithm to become one of the CO2M retrieval algorithms operated in EUMETSAT’s ground segment. Within our presentation, we will discuss the most recent retrieval developments incorporated in FOCAL-OCO2 v10 and present the corresponding improved XCO2 data product which is part of ESA’s GHG-CCI+ climate research data package 7 (CRDP7). The retrieval developments comprise a new cloud filtering technique by means of a random forest classifier, usage of a new CO2 a priori climatology, a new bias correction scheme using a random forest regressor, modifications of the radiative transfer, and others. The improved global data product exhibits an about three times higher data density and spans a time period of eight years (2014-2021). The results of a validation study using TCCON data will also be presented.

M.Reuter, M.Buchwitz, O.Schneising, S.Noel, V.Rozanov, H.Bovensmann and J.P.Burrows: A Fast Atmospheric Trace Gas Retrieval for Hyperspectral Instruments Approximating Multiple Scattering - Part 1: Radiative Transfer and a Potential OCO-2 XCO2 Retrieval Setup Remote Sensing, 9(11), 1159; doi:10.3390/rs9111159, 2017a

M.Reuter, M.Buchwitz, O.Schneising, S.Noel, H.Bovensmann and J.P.Burrows: A Fast Atmospheric Trace Gas Retrieval for Hyperspectral Instruments Approximating Multiple Scattering - Part 2: Application to XCO2 Retrievals from OCO-2 Remote Sensing, 9(11), 1102; doi:10.3390/rs9111102, 2017b

Noël, S., Reuter, M., Buchwitz, M., Borchardt, J., Hilker, M., Bovensmann, H., Burrows, J. P., Di Noia, A., Suto, H., Yoshida, Y., Buschmann, M., Deutscher, N. M., Feist, D. G., Griffith, D. W. T., Hase, F., Kivi, R., Morino, I., Notholt, J., Ohyama, H., Petri, C., Podolske, J. R., Pollard, D. F., Sha, M. K., Shiomi, K., Sussmann, R., Té, Y., Velazco, V. A., and Warneke, T.: XCO2 retrieval for GOSAT and GOSAT-2 based on the FOCAL algorithm Atmospheric Measurement Techniques, 14, 3837–3869, doi:10.5194/amt-14-3837-2021, 2021

To support the ambition of national and EU legislators to substantially lower greenhouse gas (GHG) emissions as ratified in the Paris Agreement on Climate Change, an observation-based "top-down" GHG monitoring system is needed to complement and support the legally binding "bottom-up" reporting in national inventories. For this purpose, the European Commission is establishing an operational anthropogenic GHG emissions Monitoring and Verification Support (MVS) capacity as part of its Copernicus Earth observation programme. A constellation of three CO2, NO2, and CH4 monitoring satellites (CO2M) will be at the core of this MVS system. The satellites, to be launched from 2026, will provide images of CO2, NO2, and CH4 at a resolution of about 2 km × 2 km along a 250-km wide swath. This will not only allow observing the large-scale distribution of the two most important GHGs (CO2 and CH4), but also capturing the plumes of individual large point sources and cities.
Emissions of point sources can be quantified from individual images using a plume detection algorithm followed by data-driven methods computing cross-sectional fluxes or fitting Gaussian plume models. To estimate annual emissions, a sufficiently large number of estimates is required to limit the uncertainty due to the temporal variability of emissions. However, the number of detectable plumes is limited, because the signal-to-noise ratio of individual plumes is too low or because neighboring plumes are overlapping. We present methods for increasing the number of plumes available for emission quantification using computer vision technqiues and improved data-driven methods that can estimate emissions from overlapping plumes.
Using synthetic data generated in the SMARTCARB project (Kuhlmann et al., 2020), we show that a joint denoising of coincident CO2 and NO2 images can result in signficantly improved signal-to-noise ratios for the individual images (notably, improving the peak signal-to-noise ratio of the CO2 images by +13 dB). Furthermore, by using a generative adverserial neural network approach, we show that it is possible to fill in missing data due to, e.g., cloud cover, with as an additional input wind direction information to steer the interpolation for the missing data. This ‘inpainting’ method helps the segmentation step, as it becomes possible to connect otherwise disjoint parts of a plume. Finally, we show how plume detection may be improved to be particularly receptive to plume-like features on satellite images (e.g., stretched out and narrow enhancements over the background) using a method referred to as Meijering.
A remaining challenge is to quantify the emissions from overlapping plumes, e.g., those occurring when one point source lies in the downwind direction of another plume, or when two diffusive plumes are positioned close to each other. We developed a data-driven approach using a multi-plume model that alleviates this problem. First, the approach obtains a best fitting center line for each of the individual plume sources, using effective wind data information and the multimodal distribution in the CO2 and NO2 images as inputs. Once such center lines are available, a cross-sectional flux method assuming a Gaussian cross-sectional structure can be computed for the multiple plume sources simultaneously. The upstream part of the plume (prior to overlapping) can be used to constrain the estimated fluxes. An alternative solution is to find best-fitting parameters for two or more Gaussian plume models simultaneously to estimate the emissions of each point source.
The improvements in the plume detection algorithm, and the multi-plume models for estimating emissions of overlapping plumes, increase the number of satellite images from which emission can be quantified. The larger number of emission estimates reduces the uncertainties in estimated annual emissions for point sources.

Methane (CH₄) is an important anthropogenic greenhouse gas and its rising concentration in the atmosphere contributes significantly to global warming. Satellite measurements of the column-averaged dry-air mole fraction of atmospheric methane, denoted as XCH₄, can be used to detect and quantify the emissions of methane sources. This is important since emissions from many methane sources have a high uncertainty and some emission sources are unknown. In addition, sufficiently accurate long-term satellite measurements provide information on emission trends and other characteristics of the sources, which can help to improve emission inventories and review policies to mitigate climate change.
The Sentinel-5 Precursor (S5P) satellite with the TROPOspheric Monitoring Instrument (TROPOMI) onboard was launched in October 2017 into a sun-synchronous orbit with an equator crossing time of 13:30. TROPOMI measures reflected solar radiation in different wavelength bands to generate various data products and combines daily global coverage with high spatial resolution. TROPOMI's observations in the shortwave infrared (SWIR) spectral range yield methane with a horizontal resolution of typically 7x7km².
We use a monthly XCH₄ data set (2018-2020) generated with the WFM-DOAS retrieval algorithm, developed at the University of Bremen, to detect locally enhanced methane concentrations originating from emission sources.
Our detection algorithm consists of several steps. At first, we apply a spatial high-pass filter to our data set to filter out the large-scale methane fluctuations. The resulting anomaly ∆XCH₄ maps show the difference of the local XCH₄ values compared to its surroundings. We then use these monthly maps to identify regions with local methane enhancements by utilizing different filter criteria, such as the number of months in which the local methane anomalies ∆XCH₄ of a possible hot spot region must exceed a certain threshold value. In the last step, we calculate some properties of the detected hot spot regions like the monthly averaged methane enhancement and attribute the hot spots to potential emission sources by comparing them with inventories of anthropogenic methane emissions.
In this presentation, the algorithm and initial results concerning the detection of local methane enhancements by spatially localized methane sources (e.g. wetlands, coal mining areas, oil and gas fields) are presented.

Anthropogenic greenhouse gas (GHG) emissions in the Eastern Mediterranean and Middle East (EMME) have increased fivefold over the last five decades. Emission rates in this region were ~3.4 GtCO2eq/yr during the 2010s, accounting for ~7% of the global anthropogenic GHG emissions. Among various GHGs emitted, methane (CH4) is of particular interest, given its stronger global warming potential relative to CO2 and the role of EMME as a key oil and gas producing region. Bottom-up inventories have reported that the anthropogenic CH4 emissions in EMME were ~22.0 Tg/yr in the 2010s, of which ~70% were contributed by oil and gas sectors. As inventory-based estimates often suffer from uncertainties in emission factors and activity statistics, independent budget estimation based on atmospheric observations, preferably at regional or national scales, are required to verify inventories and evaluate effectiveness of climate mitigation measures. Meanwhile, the availability of satellite CH4 observations in the recent decade (notably GOSAT XCH4 and TROPOMI XCH4) provides new opportunities to constrain CH4 emissions in this region previously underrepresented by ground-based observational networks. Here, we present a study of CH4 inverse modeling over EMME, using a Bayesian variational inversion system PYVAR-LMDz-SACS developed by LSCE, France with satellite XCH4 observations. The inversion system takes advantage of the dense XCH4 observations from space and the zooming capability of the atmospheric transport model LMDz to resolve CH4 emissions in EMME at a spatial resolution of ~50km. Instead of the default model settings for global CH4 inversions, we adapt the definition of error structure in the inversion system wherever necessary to address issues with ultra-emitters (which are common in the study region) at the high spatial resolution. The inversion results are evaluated against independent observations within and outside the study region from various platforms, and compared with emission inventories and other global or regional inversion products. With these datasets and modeling tools, we aim to assess the variations in CH4 emissions in EMME at the scales where decision-making and climate actions take place.

Globally, the oil, gas and coal sectors are the main emitters of anthropogenic methane (CH4) from fossil fuel sources. Together, these sectors represent one third of total global anthropogenic CH4 emissions. Despite a reduction in some basins due to the COVID-19 pandemic, global emissions from the oil and gas sector have rapidly increased over the last decades. This study presents results of global atmospheric inversions and compares them to national reports and other emission inventories. Results show that inversions tend to estimate higher CH4 emissions compared to national reports of oil-and-gas-producing countries like Russia, Kazakhstan, Turkmenistan and those located in the Arabic Peninsula. This difference might be partially explained since ultra-emitting events, consisting of large and sporadic emissions (greater than ≈ 20 tCH4 per hour), are not considered by emission inventories. Ultra-emitters are especially important in some countries, such as Kazakhstan and Turkmenistan, where estimated ultra-emitter emissions are comparable (1.4 Tg yr-1) to total fossil fuel emissions reported in their national inventories (1.5 Tg yr-1) and to half (on average) of the values reported in the other inventories that we analyzed. This study also considers emissions derived from regional inversions using S5P-TROPOMI atmospheric measurements at the scale of regional extraction basins for oil, gas and coal. Here, we assumed that those basins are already counted as part of the national CH4 budgets from in-situ-driven and GOSAT-driven inversions. Two coal basins, one in the USA and one in Australia, were considered. Also, six major oil and gas basins (3 in the USA, 2 in the Arabian Peninsula, and 1 in Iran) were considered as specific areas where many individual wells and storage facilities are concentrated. Averaged emissions (2019-2020) from the Bowen basin in Australia are greater than 2017 emissions estimated by inversions. For the USA, emissions from all basins analyzed, account for ~60% of total USA fossil fuel emissions estimated by inversions. For oil and gas, a basin encompassing four of the highest oil-producing fields in the world (comprising Iraq and Kuwait) represents ~38% of the total fossil emissions estimated by inversions for the Arabian Peninsula. Lastly, the basin estimation for Iran (2.5 TgCH4) represents ~68% of fossil fuel emissions from inversions and ~59% of independent inventories. Given the important role of the oil, gas and coal sectors to global anthropogenic emissions of CH4, our synthesis allows interpreting the main apparent differences between a large suite of recent emission estimates for these sectors.

The RAL Remote Sensing Group has developed an optimal estimation scheme to retrieve global height-resolved information on methane from IASI using the 7.9 µm band. This scheme uses pre-retrieved temperature, water vapour and surface spectral emissivity from the RAL Infrared Microwave Sounder (IMS) retrieval scheme, based on collocated data from IASI, MHS and AMSU. The IASI methane retrieval scheme has been used to reprocess the IASI MetOp-A record, producing a global 10-year v2.0 dataset (2007-17) (http://dx.doi.org/10.5285/f717a8ea622f495397f4e76f777349d1) and has also been applied to IASI on MetOp-B to extend the record to 2021.

While providing information on two independent vertical layers in the troposphere, sensitivity in the 7.9 µm band decreases towards the ground, due to decreasing thermal contrast between the atmosphere and surface. A combined scheme exploiting the high signal-to-noise information from Sentinel-5P (SWIR/column) with that from IASI MetOp-B (TIR/height-resolved) would enable lower tropospheric distributions of methane to be resolved. Lower tropospheric concentrations are more closely related to emission sources than are column measurements and inverse modelling of surface fluxes should be less sensitive to errors in representation of transport at higher altitudes; a limiting factor for current schemes.

Here we present findings from the IASI methane v2.0 dataset and introduce the RAL SWIR-TIR scheme, which combines Level 2 products from Sentinel 5P and IASI/CrIS to resolve lower tropospheric methane and carbon monoxide.

The Arctic and boreal regions have unique and poorly understood natural carbon cycles as well as increasing anthropogenic activities, e.g. from the oil and gas industry sector. The evolution of the high-latitude carbon sources and sinks would be most comprehensively observed by satelllites, in particular the planned Copernicus Anthropogenic CO2 Monitoring mission (CO2M). However, high latitudes pose significant challenges to reliable space-based observations of greenhouse gases. In addition to large solar zenith angles and frequent cloud coverage, snow-covered surfaces absorb strongly in the near-infrared wavelengths. Because of the resulting low radiances of the reflection measured by the satellite in nadir geometry, the retrievals over snow may be less reliable and are, for existing missions, typically filtered or flagged for potentially poor quality.

Snow surfaces are highly forward-scattering and therefore the traditional nadir-viewing geometries over land might not be optimal and instead the strongest signal could be attainable in glint-like geometries. In addition, the contributions from the 1.6 um and 2.0 um CO2 absorption bands need to be evaluated over snow. In this work, we examine the effects of a realistic, non-Lambertian surface reflection model of snow based on snow reflectance measurements on simulated top-of-atmosphere radiances in the wavelength bands of interest. The radiance simulations were carried out with various different viewing geometries, solar angles and snow surfaces. The effect of off-glint pointing was also investigated.

There are three main findings of the simulation study. Firstly, snow reflectivity varies greatly by snow type, but the forward reflection peak is present in all examined types. Secondly, glint observation mode was found to be more reflective than nadir observation mode over snow surfaces across all the examined wavelengths bands and geometries. Thirdly, the weak CO2 band had systematically greater radiances than the strong CO2 band which could indicate a greater significance in retrievals over snow.

ESA SNOWITE is a feasibility study funded by European Space Agency for examining how to improve satellite-based remote sensing of CO2 over snow-covered surfaces. It is a cooperative project between Finnish Meteorological Institute, Finnish Geospatial Research Institute and University of Leicester. The primary aim of the project is to support the development of the planned CO2M mission.

Satellite observations of greenhouse gases (GHG) are greatly enhanced when used in conjunction with ground-based sensor networks. By using clusters of spectroscopic instruments measuring GHG column abundances at locations along the satellite overpass, critical validation data for satellite GHG measurements can be provided.

In particular, missions such as NASA OCO-3 and the upcoming UKSA-CNES MicroCarb – which will provide measurements of CO2 over cities – would be aided by the presence of such ground based networks around and within urban areas. These would act as both validation sites for satellite GHG measurements, and as a long term measurement network of GHG column abundances, improving the understanding of carbon dynamics within urban areas.

However, there has previously existed a gap in the provision of such ground based networks, due to expense, infrastructure concerns, and the ability to provide autonomously acquired, high resolution data. This issue is exacerbated in areas of restricted or minimal site infrastructure, such as in city centres or remote site locations of interest e.g. peatlands, tropical forests. To fill this gap, the NERC Field Spectroscopy Facility (FSF) has developed the Spectral Atmospheric Suite (SAS), a suite of high resolution, portable and autonomous spectroscopic instruments which can be deployed by FSF as a cluster network, available for research communities in the UK and internationally.

The SAS consists of three discrete instrument “nodes”, which can be deployed individually or as part of a network cluster. Each instrument node consists of a Fourier Transform Infrared (FTIR) spectrometer (the EM27/SUN (Bruker GmbH, Germany), spectral range: 5,000 – 14,500 cm-1), measuring the column abundances of CO2, CH4 and CO; a 2D MAX-DOAS (the 2D SkySpec (AirYX GmbH, Germany), spectral range: 300 – 565 nm), measuring the slant column densities of a range of trace gases including NO2 and SO2; an automatic weather station (Vaisala, Finland) measuring meteorological parameters required for the retrievals of GHGs and trace gases; and a sun-sky-lunar sunphotometer (CIMEL Electronic, France), measuring aerosol optical thickness. Combined, each node represents an autonomous “miniature supersite”, capable of providing long term measurements as a ground based validation site for satellite measurements of GHGs and other trace gases. Each node is portable and has a low spatial footprint, allowing for easy deployment in areas of minimal or restricted infrastructure, such as city centres or remote wetland regions.

We present here an overview of the NERC FSF Spectral Atmospheric Suite and how, as part of its current deployment until 2022 with the University of Leicester’s London Carbon Emissions Experiment, it will provide a ground based validation site for upcoming missions such as UKSA-CNES MicroCarb.

The emissions of halocarbons have profoundly modified the chemical and radiative equilibrium of our atmosphere. These halogenated compounds are known to be powerful greenhouse gases and contribute, for chlorinated and fluorinated compounds, to the depletion of stratospheric ozone and to the development of the ozone hole. Their monitoring is therefore essential. The aim of this work is to assess the potential of infrared satellite sounders operating in the nadir geometry, to contribute to this monitoring and thereby to complement existing surface measurement networks.
This work is centered on the exploitation of the measurements from the infrared satellite sounder IASI. The instrument stability and the consistency between the different instruments on the successive Metop platforms (A, B and C) is remarkable and makes it a reference for climate monitoring. Among other things, IASI offers the potential to investigate trends in the atmospheric abundance of various species better than with any other hyperspectral IR sounder. The low noise of the IASI radiances is also such that even weakly absorbing halocarbons can be identified. Recently, we managed to detect the spectral signatures of eight halocarbons: CFC-11, CFC-12, HCFC-22, HCFC-142b, HFC-134a, CF4, SF6 and CCl4. In this work we exploit the 15 years record of continuous IASI measurements to give a first assessment of the trend evolution of these species. This is done by targeting various geographical areas on the globe and examining the remote oceanic and continental source regions separately. The trend evolution in the different chemical species, either negative or positive, is validated against what is observed with ground-based measurement networks and other remote sensors. We conclude by assessing the usefulness of IASI and follow-on missions to contribute to the global monitoring of halocarbons.

The paper presents the results of the CarbonCGI study proposed by ESA and carried out by Thales Alenia Space and partners for the observation of faint GHG source’s emissions with a high resolution Compact Gas Imager (CGI).
Atmospheric remote sensing from CGI allows observation of atmosphere features ranging from largest scales of meteorology and smaller spatial scales down to the finest scales allowing direct observations of biogenic and anthropogenic inter-actions with atmosphere. For this, multi-mission deployment is foreseen from geostationary to low orbit satellites as well as on airborne platform and on ground for mobile applications. CGI has the potential to acquire high resolution images of gas in the spectral regions of solar emission from UV to SWIR, and also to take images of atmospheric temperature and humidity profiles in TIR spectral bands.
This paper is focused on the detection and characterisation of Carbon dioxide and Methane gas concentrations, from low orbit satellite, for climate applications. CarbonCGI development includes simulation and experimental validation of level 0 (instrument design and acquisition chain), level 1 (data correction), level 2 (Radiative Transfer Model), and level 4 (Transport Model). CarbonCGI is developed in an integrated team of scientists and engineers. Knowledge in the field of physics of atmosphere from laboratories and scientific engineering institutes aim at designing the most efficient atmosphere remote sensor. Thus, the described CGI principle optimises the retrieval of atmospheric states from the spectra variability with acquisition of specific Partially Scanned Interferograms (PSI), resulting from a double optimisation of both spectral bands and Optical Path Difference range. The optical concept works at low aperture number, and provides very long dwell time, to reach unprecedented radiometric resolution with very low sounding precision and accuracy, in a very high spatial resolution image.
The paper presents the results obtained by applying the Performance Simulation Platform developed in the framework of the scientific chair TRACE https://trace.lsce.ipsl.fr to the CarbonCGI imaging and sounding performance. The obtained results highlight the capacity to carry out early mission trade-off from acquisition chain parameters.
The sounding performance obtained by coupling level 0-1 and level 2 models are described. After the presentation of level 0-2 models, this paper presents the sounding performance which have been achieved during the optimisation of the acquisition chain design. CGI instrument design delivers an inherent solution to correct for the presence of atmospheric aerosol up aerosol optical depths of 1. An optimised aerosol’s bias measurement concept and associated models and performance are presented.
Level 0-1 activity are then summarised with the presentation of the payload’s design, optical thermal and mechanical design, with an introduction to the CarbonCGI stray light model and the CarbonCGI Line of Sight Stabilisation system inferred from the ISABELA LS3 design, developed in the frame of the ISABELA TRP for ESA.
The paper concludes with a proposal of an incremental implementation plan of weak source measurement missions based on high resolution CarbonCGI imagers. The first step is a CarbonCGI instrument to complement CO2M mission with observation at higher spatial resolution and smaller swath, the second step is a self-standing high resolution observing system.

Greenhouse gas measurements by a Fourier Transform Spectrometer (FTS) were established at Sodankylä (67.4° N, 26.6° E) in early 2009 (Kivi and Heikkinen, 2016). The instrument records high-resolution solar spectra in the near-infrared spectral region. From the spectra we derive column-averaged, dry-air mole fractions of methane (XCH4), carbon dioxide (XCO2) and other gases. The instrument participates in the Total Carbon Column Observing Network (TCCON). Sodankylä is currently the only TCCON site in the Fennoscandia region. Our measurements have contributed to the validation of several satellite-based instruments. The relevant satellite missions include Sentinel-5 Precursor by ESA, the Orbiting Carbon Observatory-2 (OCO-2) by NASA (e.g., Wunch et al., 2017), the Greenhouse Gases Observing Satellite (GOSAT/GOSAT-2) mission by JAXA and the Chinese Carbon Dioxide Observation Satellite Mission (TanSat).
Comparisons with the GOSAT observations of XCH4 and XCO2 taken during years 2009-2020 show good agreement. Mean relative difference in XCH4 has been 0.04 ± 0.02 % and the mean relative difference in XCO2 has been 0.04 ± 0.01 %. We also performed a series of AirCore flights during each season in order to compare FTS retrieval results with the AirCore measurements. AirCore sampling system is directly related to the World Meteorological Organization in situ trace gas measurement scales. Thus the AirCore data can be used to calibrate remote sensing instruments. Our AirCore is a 100 m long coiled sampling tube with a volume of approximately 1400 ml. The sampler is lifted by a meteorological balloon typically up to about 30-35 km altitude and is filled during descent of the instrument from stratosphere down to the Earth's surface. Shortly after landing we have analyzed the sample using a cavity ring-down spectrometer. In addition to the balloon borne AirCore flights we also took measurements of methane and carbon dioxide at a 50 meter tower and by a drone borne AirCore instrument in the vicinity of the FTS site.

Kivi, R. and Heikkinen, P.: Fourier transform spectrometer measurements of column CO2 at Sodankylä, Finland, Geosci. Instrum. Method. Data Syst., 5, 271–279, https://doi.org/10.5194/gi-5-271-2016, 2016.
Wunch, D., et al., Comparisons of the Orbiting Carbon Observatory-2 (OCO-2) XCO2 measurements with TCCON, Atmos. Meas. Tech., 10, 2209-2238, https://doi.org/10.5194/amt-10-2209-2017, 2017.

The Arctic Observing Mission (AOM) is a satellite mission concept that would use a highly elliptical orbit (HEO) to enable frequent observations of greenhouse gases (GHGs), air quality, meteorological variables and space weather to address the current sparsity in spatial and temporal coverage north of the usable viewing range of geostationary (GEO) satellites. AOM evolved from the Atmospheric Imaging Mission for Northern Regions (AIM-North), which was expanded in scope. AOM would use an Imaging Fourier Transform Spectrometer (IFTS) with 4 near infrared/shortwave infrared (NIR/SWIR) bands to observe hourly CO2, CH4, CO and Solar Induced Fluorescence spanning cloud-free land from ~40-80°N. The rapid revisit is only possible due to cloud avoidance using ‘intelligent pointing’, which is facilitated by the availability of real-time cloud data from the meteorological imager and the IFTS scanning approach. Simulations suggest that these observations would improve our ability to detect and monitor changes in the Arctic and boreal carbon cycle, including CO2 and CH4 emissions from permafrost thaw, or changes to northern vegetation carbon fluxes under a changing climate. AOM is envisioned as a Canadian-led mission to be implemented with international partners. AOM is currently undergoing a pre-formulation study to refine options for the mission architecture and advance other technical and design aspects, investigate socio-economic benefits of the mission and better establish the roles and contributions of partners. This presentation will give an overview of the AOM GHG instrument, its expected capabilities and its potential for carbon cycle science and monitoring.

A large and fast view of volcanic plumes as detection and measurement of volatiles components exolving from craters is possible by using hyperspectral remote sensing if their absorption bands are in the sensor spectral range. In the present study the developed algorithm to calculate CO2 columnar abundance in tropospheric volcanic plume is presented. The algorithm is based on a modified CIBR 'Continuum Interpolated Band Ratio' remote sensing technique based on a differential absorption technique that was initially developed to calculate water vapor columnar abundance. The retrieval techniques exploit spectroscopy measurements by analysing gases absorptions features in the SWIR (Short Wave InfraRed) spectral range, in particular the Carbon Dioxide absorption in the spectral range of 2 microns. Specifically, PRISMA (PRecursore IperSpettrale della Missione Applicativa) acquisition data are used for gases retrieval purposes. The PRISMA space mission was launched by the Italian Space Agency (ASI) on March 22, 2019; the on-board spectrometers are able to measure in two spectral range: the VNIR (0.4-1.0 µm) and SWIR (0.9-2.5 µm) spectral ranges, with a ground spatial resolution of 30 m. In this study, the inversion techniques is applied to PRISMA data in order to derive the PRISMA performances for CO2 detection and retrieval. Simulations of the “Top Of Atmosphere (TOA)” radiance have been performed by using real input data to reproduce the scene acquired by PRISMA over a volcanic point sources: actual atmospheric background of CO2 (~400 ppm) and vertical atmospheric profiles of pressure, temperature and humidity obtained from probe balloons has been used in the radiative transfer model. The results will be shown in the considered test sites of Campi Flegrei caldera in the Campania region (located in southern Italy) and Lusi volcanoes (located in Java island region of Indonesia) both characterized by a persistent degassing plume present even if they show a very different mechanism of volcanic emissions: the first based on a hydrothermal system and the second based on a mud cold mechanism of volcanic emission of gases in troposphere.

Satellite observations of carbon dioxide have recently matured to the level where they can be used to estimate anthropogenic CO2 emissions of large power plants and other point sources. The true added value of these observations is gained specifically over regions that are otherwise not measured or where the reported emission inventories may be defective. However, satellite observations of CO2 are sensitive to other atmospheric pollutants, specifically aerosol particles that affect the path length of radiation through scattering and absorption. For further complication, these particles are often co-emitted with anthropogenic CO2 emissions. The impact of aerosols on CO2 retrievals can be considered to some extent in the retrieval process and post-processing bias correction. Still, little attention has been dedicated to the evaluation of CO2 retrievals in high aerosol loadings that are characteristic to megacity environments and other regions with persistently poor air quality and high aerosol optical depth (AOD).

In this work we present two approaches to investigate potential aerosol effects on OCO-2 XCO2 observations. To obtain global statistics, a co-located database for OCO-2 XCO2 (OCO-2 v10r) and MODIS Aqua AOD (L2, 10km Dark Target) is created. For each OCO-2 pixel the corresponding MODIS AOD value was defined from the nearest good quality MODIS observation that was found within 0.2 deg. lat., lon. distance from the XCO2 observation. The dataset consists of 5 years of observations between 2015 and 2019. This unique global dataset enables the investigation of large scale variation patterns, regional dependencies and allow also to identify potentially interesting areas for more detailed study. In the local scale approach, the aerosol effects are studied in the vicinity of urban TCCON stations, that also have an operating AERONET or other sunphotometer station close by. To investigate the spatial patterns, L2 MODIS Aqua 3km Dark Target AOD is analysed together with L2 OCO-2 XCO2. Hence, in this approach for both XCO2 and AOD, a ground-based reference measurement can be obtained in addition to the satellite observations. Also, aerosol vertical profiles from Calipso will be analysed if an overpass over the study area is obtained. With this combination of observations the potential risks for aerosol induced biases in a city scale can be assessed in detail.

This research will lay important ground work to the planned Copernicus Anthropogenic CO2 Monitoring Mission where the ultimate purpose is to support the goals of the Paris Agreement with independent emission estimates derived from satellite observations. For this purpose, it is crucial to investigate the validation of CO2 observations in urban, high AOD environments and establish the current state of the art and gaps in both retrievals and validation.

Methane is the most important anthropogenic greenhouse gas after carbon dioxide. In fact, it is responsable for about one quarter of the climate warming experienced since preindustrial times. A considerable amount of these emissions comes from methane point-sources, typically linked to fuel production installations. Thus, detection and elimination of these emissions represents a key means to reduce the concentration of greenhouse gases in the atmosphere.

A functional global monitoring of methane emissions is possible because of satellites, which capture the upwelling radiance at the top-of-atmosphere level in different spectral bands. One example of this technology is the Sentinel-5P TROPOMI mission which monitors methane emissions at a global scale and daily revisit. However, its relatively low spatial resolution cannot pinpoint methane point-source emissions with high accuracy. In contrast, the Italian PRISMA mission presents a lower temporal revisit but a larger spatial resolution of 30 m and measures the top-of-atmosphere radiance in the 400–2500 nm spectral range, where significant methane absorption features are well characterized. Therefore, PRISMA mission can largely complement the capabilities of TROPOMI for the detection and quantification of methane at a global scale.

In this study, different methodologies for point-source methane retrieval detection and quantification using PRISMA data have been reviewed in order to determine the most accurate procedure. The review goes from multitemporal methods that compare data from different days with methane emission to days with no emission to target detection algorithms such as the simple matched-filter based algorithm applied to the ~2300 nm methane absorption window in shortwave infrared spectral region. The accuracy of the different methodologies has been assessed under different scenarios that consider the most relevant error sources in the retrieval such as the surface brightness and homogeneity. This assessment has flagged the main areas of potential improvement of the retrieval methodologies and, consequently, several techniques have been developed that include the detection of false positives (e.g. the identification of plastic and hydrocarbons) and the minimisation of the surface heterogeneity impact.

In recent years at ECMWF, a series of projects were carried out focusing on developments towards direct assimilation and monitoring to exploit space-borne cloud radar and lidar data for Numerical Weather Prediction (NWP) models. Although active observations from such profiling instruments contain a wealth of information on the structure of clouds and precipitation, they have never been assimilated directly in any global NWP model.

To prepare the data assimilation system for the new observations of cloud radar reflectivity and lidar backscatter, several important developments were required. This included the specification of sufficiently accurate observation operators (i.e. models providing equivalent model fields to observations), as well as defining flow-dependent observation errors, and appropriate quality control strategy and bias correction scheme. The feasibility of assimilating CloudSat and CALIPSO data, currently the only available data from space-borne radar and lidar with global coverage, into the Four-Dimensional Variational (4D-Var) data assimilation system used at ECMWF has been investigated. Including cloud radar reflectivity and lidar backscatter in the assimilation system had a positive impact on both the analysis and the subsequent short-term forecast. By running experiments for different seasons and combining them to increase statistical significance lead to promising results; improvements to the zonal mean forecast skill score in the short- and medium-ranges for large-scale variables were found almost anywhere, with the largest impact on storm-tracks and in the tropics.

The performed studies using CloudSat and CALIPSO observations prepared grounds for assimilation of such observation types from the future EarthCARE mission. Additionally, the system developments will facilitate the monitoring of observations both in an operational sense and for model evaluation as soon as observations become available after the mission launch. By using a monitoring system that combines information from observations and model, a statistically significant drift in the measurements can be detected faster than monitoring observations alone. Also the monitoring system allows validation of the observations along the whole EarthCARE track.

Daytime Polarization Calibration Using Solar Background Signal Scattered from Dense Cirrus Clouds in the Visible and Ultraviolet Wavelength Regime
Zhaoyan Liu, Pengwang Zhai, Shan Zeng, Mark Vaughan, Sharon Rodier, Xiaomei Lu, Yongxiang Hu, Charles Trepte, and David Winker

In this presentation we describe the application of a previously developed technique that is now being used to correct the daytime polarization calibration of the CALIPSO lidar [1]. The technique leverages the fact that the CALIOP solar radiation background signals measured above dense cirrus clouds are largely unpolarized [2] due to the internal multiple reflections within the non-spherical ice particles and the multiple scattering that occurs among these particles. Therefore, the ratio of polarization components of the cirrus background signals provides a good estimate for the polarization gain ratio (PGR) of the lidar. Using airborne backscatter lidar measurements, this technique was demonstrated to work well in the infrared regime where the contribution from the molecular scattering between dense clouds is negligible. However, in the visible and ultraviolet regime, the molecular contribution is too large to be ignored, and thus corrections must be applied to account for the highly polarizing characteristics of the molecular scattering. Ignoring these molecular scattering contributions can cause PGR errors of 2-3% at 532 nm, where the CALIPSO lidar makes its depolarization measurement. Because of the wavelength dependence of -4 of the molecular scattering, the PGR error can be even larger at the 355 nm wavelength that will be used by ESA’s EarthCARE lidar. To estimate the molecular scattering contributions to the lidar received solar background signal, a look-up table has been created using a polarization-sensitive radiative transfer model [3]. This presentation describes the theory and implementation of the molecular scattering correction, demonstrates the application of the calibration technique, and compares the results to CALIOP daytime PGR estimates derived using an onboard pseudo-depolarizer [4]. We also present the simulation results at 355 nm at the symposium.
References:
1. Z. Liu, M. McGill, Y. Hu, C. Hostetler, M. Vaughan, and D. Winker, “Validating lidar depolarization calibration using solar radiation scattered by ice clouds”, IEEE Geos. and Remote Sensing Lett., 1, 157-161, 2004.
2. K. N. Liou, Y. Takano, and P. Yang et al., “Light scattering and radiative transfer in ice crystal clouds: applications to climate research,” in Light Scattering by Nonspherical Particles, M. Mishchenko et al., Eds. San Diego, CA: Academic, 2000, pp. 417–449.
3. P. Zhai, Y. Hu, J. Chowdhary, C. R. Trepte, P. L. Lucker, D. B. Josset, “A vector radiative transfer model for coupled atmosphere and ocean systems with a rough interface”, Journal of Quantitative Spectroscopy and Radiative Transfer, 111, 1025-1040, 2010.
4. J. P. McGuire and R. A. Chapman, “Analysis of spatial pseudo depolarizers in imaging systems,” Opt. Eng., vol. 29, pp. 1478–1484, 1990.

The Earth Cloud, Aerosols and Radiation Explorer (EarthCARE) is a joint mission of the European Space Agency (ESA) and the Japan Aerospace Exploration Agency (JAXA). The mission objectives are to improve the understanding of the cloud-aerosol-radiation interactions by acquiring vertical profiles of clouds and aerosols simultaneously with radiance and flux observations for their better representation in numerical atmospheric models
The operational EarthCARE L2 product on top-of-atmosphere (TOA) radiative fluxes is based on a radiance-to-flux conversion algorithm fed mainly by unfiltered broad-band radiances from the BBR instrument, and auxiliary data from EarthCARE L2 cloud products and modelled geophysical databases. The conversion algorithm models the angular distribution of the reflected solar radiation and thermal radiation emitted by the Earth-Atmosphere system, and returns flux estimates to be used for the radiative closure assessment of the Mission.

Different methods are employed for the solar and thermal BBR flux retrieval models. Models for SW radiances are created for different scene types and constructed from Clouds and the Earth’s Radiant Energy System (CERES) data using a feed-forward back-propagation artificial neural network (ANN) technique. LW models are based on correlations between BBR radiance field anisotropy and the spectral information provided by the narrow-band radiances of the imager instrument on-board. Both retrieval algorithms exploit the multi-viewing capability of the BBR (forward, nadir and backward observations of the same target) co-registering radiances and providing flux estimates for every view and checking their integrity before being combined into the optimal flux of the observed target. The reference height where the three BBR measurements are co-registered corresponds to the height where most reflection or emission takes place and depends on the spectral regime. LW observations are co-registered at the cloud top height, but the most radiatively significant height level on SW radiances is very dependent on the cloud. This reference height is instead selected by minimizing the flux differences between nadir, fore and aft fluxes.

The study presented here shows an evaluation of the BBR radiance-to-flux conversion algorithms using scenes from the Environment Canada and Climate Change’s Global Environmental Multiscale (GEM) model. The EarthCARE L2 team has simulated three EarthCARE frames (1/8 of orbit) running a radiative transfer code optimized for the EarthCARE instrument models over the GEM scenes. The test scenes resulting include synthetic L1 EarthCARE data that have been used by the different L2 teams to test and develop L2 products for testing the end-to-end-processor chaining. The test scenes collect data for a ~6200 x 150 km swath, with 1 km along-track sampling, of a simulated EarthCARE orbit. The “Halifax” scene corresponds to an orbit crossing the Atlantic Ocean and Canada in December 7, 2015. This case includes Sun just below the horizon over Greenland, cold air over Labrador, a cold-front near Halifax, dense overcast south of Halifax, and scattered shallow convection south of Bermuda. The “Baja” scene corresponds to an orbit crossing Canada and USA in April 2, 2015. This case includes clear and cold conditions at the northern extremity, scattered cloud through the Canadian Prairies, overcast over the Rocky Mountains, clear through Utah, and cirrus in Arizona and Mexico. The “Hawaii” scene corresponds to an orbit going through the Pacific Ocean and passing near the Hawaiian Islands in June 23rd 2014.

The BBR solar and thermal flux retrieval algorithms were successfully employed to retrieve radiative fluxes over the test scenes. The approach followed to evaluate the flux retrieval algorithms includes both testing the model performance with L2 products directly derived from the geophysical properties included in the GEM simulations (ideal case, no dependence on L2 retrievals analysed) and testing the model performance with L2 products derived from the EarthCARE L2 cloud and radiance processors (operational case, dependence on L2 Lidar and Imager cloud algorithms and L2 radiance unfiltering algorithm analysed). These two exercises allow evaluating discrepancies between retrieved and simulated fluxes, and assessing the sensitivity of the flux retrieval models to uncertainties in the cloud and radiance retrievals over a huge variety of realistic samples in three different scenes.

Clouds warm the surface in the longwave (LW) and this warming effect can be quantified through the surface LW cloud radiative effect (CRE). The global surface LW CRE is estimated using long-term observations from space-based radiometers (2000–2021) but has some bias over continents and icy surfaces. It is also estimated globally using the combination of radar, lidar and space-based radiometer over the 5–year period ending in 2011. To develop a more reliable long time series of surface LW CRE over continental and icy surfaces, we propose new estimates of the global surface LW CRE from space-based lidar observations. We show from 1D atmospheric column radiative transfer calculations, that surface LW CRE linearly decreases with increasing cloud altitude. These computations allow us to establish simple relationships between surface LW CRE, and five cloud properties that are well observed by the CALIPSO space-based lidar: opaque cloud cover and altitude, and thin cloud cover, altitude, and emissivity. We use these relationships to retrieve the surface LW CRE at global scale over the 2008–2020 time period (27 Wm-2). We evaluate this new surface LW CRE product by comparing it to existing satellite-derived products globally on instantaneous collocated data at footprint scale and on global averages, as well as to ground-based observations at specific locations. Our estimate appears to be an improvement over others as it appropriately capture the surface LW CRE annual variability over bright polar surfaces and it provides a dataset of more than 13 years long.
After presenting the principle of the algorithm used to retrieve the surface LW CRE from lidar observations only and the validation of the retrieval, we will 1) describe the modification needed for this CALIPSO algorithm to run on ATLID/EarthCare Level 1 data 2) explain the complementarity of this lidar-only surface LW CRE estimate with ATLID L2 ESA radiative product 3) show examples of science applications using the product build from CALIPSO data and describe the benefit for science of extending this record by applying this algorithm to ATLID Level 1 data.

The ESA cloud, aerosol and radiation mission EarthCARE will provide active profiling and passive imaging measurements from a single satellite platform. This will make it possible to extend the products obtained from the combined active/passive observations along the ground track into the swath by means of active/passive sensor synergy, to estimate the 3D fields of clouds and to assess radiative closure. The backscatter lidar (ATLID) and cloud profiling radar (CPR) will provide vertical profiles of cloud and aerosol parameters with high spatial resolution. Complementing these active measurements, the passive multi-spectral imager (MSI) delivers visible and infrared images for a swath width of 150 km and a pixel size of 500 m. MSI observations will be used to extend the spatially limited along-track coverage of products obtained from the active sensors into the across-track direction. In order to support algorithm development and to quantify the effect of different instrument configurations on the mission performance, an instrument simulator (ECSIM) has been developed for the EarthCARE mission. ECSIM is an end-to-end simulator capable to simulate all four instruments for complex realistic scenes. Specific ECSIM test scenes have been created from weather forecast model output data. The 6000 km long frames include clouds over the Greenland ice sheet, followed by high optical thick clouds, a high ice cloud regime as well as low-level cumulus cloud embedded in a marine aerosol layer below and an elevated intense dust layer above. These synthetic scenes make it possible to evaluate and intercompare the different cloud properties from active and passive sensors such as cloud liquid water path or cloud effective radius. Further the input of the synthetic scenes offer the opportunity to extract the extinction profiles for each MSI pixel, and to contrast them to the retrieved cloud properties and types. This approach can be used to better understand and quantify the differences between the retrieved cloud properties based on the different measurement principles (passive and active). For example, the cloud top height retrieved from MSI is an effective height of infrared emission located within the cloud, and it is important to quantify differences to the geometric cloud top height to constrain the longwave cloud radiative effect. Another quantity of interest is the cloud effective radius from CPR, which is most sensitive to large particles in the clouds, while MSI is only sensitive to very small particles on the top of the cloud. The goal is to understand the differences of the cloud products from CPR, ATLID and MSI by comparison to the reference fields to enable a consistent comparison.

The EarthCARE (Earth Clouds, Aerosol and Radiation Explorer) mission will be equipped with four co-located instruments (three from ESA and one provided by JAXA) to derived information related to aerosols/clouds/radiation and their interaction through the processing of single instrument data as well as synergistic products.
The Payload Data Ground Segment (PDGS) is the component of the overall EarthCARE Ground Segment in charge of receiving the housekeeping telemetry and instrument source packets via X-Band from the satellite, processing the packets in order to generate different product levels and disseminating them to users in few hours from sensing. The main products include level 0 (corrected and time sorted packets), level 1B (instrument science data calibrated and geolocated), level 2A (geophysical parameters derived from a single instrument) and level 2B or synergistic products (geophysical parameters derived by merging information of several EarthCARE instruments). The PDGS is also in charge of the routine calibration and monitoring of the three ESA instruments, of products quality control as well as planning of payload operations The EarthCARE PDGS consist of several components called facilities. Although the general architecture is similar to other PDGS developed by ESA for Earth Observation Missions, some evolutions were required to take into account some EarthCARE specific aspects.. In particular, the synergistic nature of the mission results in a complex processing model which involves about 30 different processors. In order to streamline the integration of this large number of processors and in anticipation of initially frequent updates, a formal modelling of the processing chain has been introduced to support automatic configuration of the processing facility. In addition, a new facility called the Level 2 TestBed has been included in the PDGS in order to allow processor developers to test their code in quasi operational conditions and in an autonomous way including the possibility to upload new processor versions without assistance from PDGS operators. The presence of a Japanese Instrument on-board also imposes tight dependencies between the ESA and JAXA components in terms of processing as well as in terms of payload planning.
This poster presents the functional and architectural breakdown of the PDGS, external interfaces including FOS (Flight Operation Segment), ECMWF and JAXA. It details the main design drivers including data latency, production model, data volumes, network bandwidth as well as interfaces with end users. The current integration status of the PDGS and its underlying facility is also presented.

The Hybrid End-To-End Aerosol Classification model (HETEAC) [1] has been developed for the upcoming EarthCARE mission [2]. This aerosol classification model is based on a combined experimental and theoretical (hybrid) approach and allows the simulation of aerosol properties, from microphysical to optical and radiative parameters of predefined aerosol types (end-to-end). In order to validate HETEAC, an aerosol typing scheme applicable to both ground-based and spaceborne lidar systems has been developed.
This novel aerosol typing scheme, based on HETEAC, applies the optimal estimation method (OEM) to a combination of lidar-derived intensive aerosol properties (i.e., concentration-independent), to determine the statistically most-likely contribution of aerosol component to the observed aerosol mixture, weighted against a priori knowledge of the system. The aerosol components considered to contribute to an aerosol mixture are four, namely fine, spherical, absorbing (FSA); fine, spherical, non-absorbing (FSNA); coarse, spherical (CS); and coarse, non-spherical (CNS). These four components have been selected from lidar-based experimental data set at 355, 532 and 1064 nm. Their optical and microphysical properties serve as a priori for the retrieval scheme and are in accordance with the ones used in the original HETEAC model, in order to ensure meaningful comparisons. In contrast to HETEAC, which is limited to observations at 355 nm only, the novel typing scheme is flexible in terms of input parameters and can be extended to other wavelengths to exploit the full potential of ground-based multiwavelength-Raman-polarization lidars and thus reduce the ambiguity in aerosol typing. It is thus an algorithm, able to be applied to EarthCARE but also to other lidar systems providing other or more optical products.
The initial guess of the aerosol components contribution that is needed to kick-of the retrieval scheme is the outcome of a decision tree. Using this initial guess, the lidar ratio (355 and 532 nm), particle linear depolarization ratio (355 and 532 nm), extinction-related Ångström exponent and backscatter-related color ratio (at the 532/1064 nm wavelength pair) are calculated (forward model). The final product is the contribution of the four aforementioned aerosol components to an aerosol mixture in terms of relative volume. Once this product meets certain quality assurance flags, it can be used to provide additional products: (a) aerosol component separated backscatter and extinction profiles, (b) aerosol optical depth per aerosol component, (c) volume concentration per component, (d) number concentration per component, (e) effective radius of the observed mixture and (f) refractive index of the mixture.
In this presentation, the aerosol typing scheme will be discussed in detail and it will be applied to several case studies. The application of the algorithm to different atmospheric load scenarios will demonstrate the algorithm’s strengths and limitations. In addition, first results between the HETEAC and OEM comparison will be presented.

References
[1] Wandinger, Ulla, et al., 2016: "HETEAC: The Aerosol Classification Model for EarthCARE." EPJ Web of Conferences. Vol. 119. EDP Sciences.
[2] llingworth, A., et al., 2014: THE EARTHCARE SATELLITE: The next step forward in global measurements of clouds, aerosols, precipitation and radiation. Bull. Am. Met. Soc., doi:10.1175/BAMS-D-12-00227.1.

The Broad-Band Radiometer (BBR) instrument on the future EarthCARE satellite (to be launched in 2023) will provide accurate outgoing solar and thermal radiances at the Top of the Atmosphere (TOA) obtained in an along-track configuration in three fixed viewing directions (fore, nadir and aft).

The BBR will measure radiances filtered by the spectral response of the instrument in two broad-band spectral channels; SW (0.25 to 4µm) and TW (0.25 to > 50µm). These radiances need to be corrected in the unfiltering process in order to reduce the effect of a limited and non-uniform spectral response of the instrument.

The unfiltering parametrization is based on a large simulated database of fine spectral resolution SW and LW radiances convolved with the spectral responses of the BBR channels. In practice, the SW and TW measurements of the BBR must be converted into solar and thermal (unfiltered) radiances. First, the LW radiance is estimated from the SW and TW measurements. Secondly, the inter-channel contaminations, i.e., the parts of the LW signal due to reflected solar radiation and of the SW signal due to planetary radiation, are accounted for. Finally, multiplicative factors are computed in order to estimate the unfiltered solar and thermal radiances from the SW and LW channels, respectively.

Regarding the algorithm, two unfiltering algorithms have been developed for the SW: stand-alone and MSI-based, and one stand-alone for the LW. The stand-alone algorithms aim to enable the unfiltering of the BBR if the MSI measurements are unavailable or degraded and it is done according to the measured broadband radiances and land use classification. The MSI-based algorithm makes use of the MSI cloud mask and cloud phase in the unfiltering process.

The study presented here shows an evaluation of the BBR unfiltered radiance estimation using the three synthetic test scenes (Halifax, Baja and Hawaii) created by the EarthCARE team from the Environment Canada and Climate Change’s Global Environmental Multiscale (GEM) model and radiative transfer data derived from them.

It is worth noting that the unfiltering is a crucial part in the BBR processing, as errors in the unfiltering will be propagated to the flux (BMA-FLX product). To this end, the unfiltering performances have been confirmed not only using the test scenes (RMSE ~ 0.5 W m-2 sr-1 for SW and LW) but also using an independent validation database for both SW and LW (RMSE < 1 Wm-2sr-1 for the SW and < 0.2 W m-2 sr-1 for the LW).

With the combination of two active instruments, a cloud radar and a high spectral resolution lidar, and a set of passive instruments, the ESA/JAXA EarthCARE mission will be the most complex satellite for aerosol, cloud and radiation measurements from space. With its so-called NARVAL payload, the German high altitude and long-range aircraft HALO is equipped with similar instruments as the upcoming satellite experiment. Having the same or similar payload on an aircraft provides the opportunity to apply and test algorithms, to investigate constraints of the future satellite mission, and to develop strategies for and perform validation studies.

Since 2013, the EarthCARE-like payload (HSRL at 532 nm with polarization sensitive channels at 532 nm and 1064 nm, Ka-band radar with 30 kW peak power, hyper-spectral radiometer, and microwave radiometer) on HALO was deployed during a number of six flight experiments and thus collected a large number of measurements that are currently used to prepare for the upcoming satellite mission. The measurements were performed at different locations including the tropical and sub-tropical North-Atlantic region up to the extra-tropical North-Atlantic and the European Mid-Latitudes. We used these measurements for comparison studies of current satellite measurements, airborne measurements and simulations, and process studies with the advantage of a much higher spatial resolution and/or sensitivity compared to the future space borne measurements. In this context, we investigated the benefit and constrains of the upcoming satellite mission and studied the effect of instrument resolution and sensitivity on the derived properties. With the combination of remote sensing measurements and airborne in-situ measurements we validated satellite retrievals by directly comparing retrieval output with measured properties. Looking ahead, we furthermore developed an elaborated proposal for an upcoming validation study addressing different locations and aspects of validation.

In this presentation we will give an overview of our EarthCARE preparation studies and their main results. We address different stages and aspects of satellite preparation; from the development of new strategies and methods, to sensitivity tests and finally towards the investigation of retrievals. By summarizing our lessons learned we will consolidate our insights which helped to shape ideas for a future validation campaign.

The synergy of radar and lidar from ground-based networks such as ARM and CloudNet/ACTRIS and the A-Train constellation of satellites has revolutionised our understanding of the global and vertical distribution of clouds and precipitation. However, while the complementary sensitivities of lidar to small ice crystals and radar to larger snowflakes can provide near-complete coverage of ice clouds and snow, the detection and vertical location of liquid cloud is much less certain. In mixed-phase, layered or precipitating cloud scenes the lidar is often quickly extinguished within the first layer, and while the radar penetrates most scenes its signal is dominated by larger precipitating hydrometeors. We use simulated EarthCARE measurements of midlatitude and tropical cloud scenes from a numerical weather model to show that these synergistic blind spots result in less than 25% of liquid clouds being detected by volume, representing only around 10% of total liquid water content.

As well as biasing global liquid cloud statistics and water budgets from spaceborne active remote sensing, these undiagnosed clouds cannot be ignored from a radiative perspective. In this study we use simulated EarthCARE measurements to evaluate the performance of EarthCARE’s synergistic retrieval of cloud and precipitation (ACM-CAP), which will assimilate a solar radiance channel from EarthCARE’s multi-spectral imager (MSI) as well as the cloud profiling radar (CPR) and atmospheric lidar (ATLID). We show that assuming that liquid clouds are collocated with precipitation improves the forward-modelled solar albedo in many complex cloud scenes. Even without active measurements of liquid cloud, the solar radiance and CPR path-integrated attenuation are sufficient to constrain the retrieval of a simplified profile of liquid water content, which reduces underestimates in retrieved liquid water path without introducing a significant compensating error. When the profiling retrievals at nadir and MSI imagery are used to reconstruct a 3D across-swath scene (ACM-3D), the missing liquid contributes to a mean bias error of almost 40 gm-2 with respect to the model fields, compared to around -5 gm-2 when liquid is included in the synergistic retrieval constrained by solar radiances. Finally the radiative closure assessment (ACMB-DF) against EarthCARE’s broadband radiometer (BBR) identifies shortwave flux deficits of 50 to 100 Wm-2 due to this undiagnosed liquid cloud associated with deep midlatitude cloud scenes, confirming that a simple assumption accounting for radar-lidar blind spots within the synergistic retrieval can result in significant improvements in retrievals of radiatively-important liquid cloud.

Validation activities are critical to ensure the quality, credibility, and integrity of Earth observation data. With the deployment of advanced active remote sensors in space, a clear need arises for establishing best practices in the field of cloud and aerosol profile validation. The upcoming EarthCARE mission brings several validation challenges arising from the multi-sensor complexity/diversity and the innovation of its standalone and synergistic products. EarthCARE is a joint ESA-JAXA mission to study interactions between clouds, aerosols, and radiation and their fundamental roles in regulating the climate system. Owing to its active remote sensing payloads, i.e. Atmospheric Lidar (ATLID) and Cloud Profiling Radar (CPR), EarthCARE is capable of performing range-resolved measurements of clouds and aerosols, which are demanding in terms of validation needs and related protocols. Furthermore, special protocols are also needed for the validation of radiance measurements from the opposite viewing direction.
With the involvement of international ground-based networks and airborne facilities in the EarthCARE validation community, there will be a wealth of correlative datasets for Cal/Val purposes. Efficient coordination will be needed between the instrument PIs (orbital and suborbital), the validation teams along with algorithm teams from related missions, and the end-user community (e.g., the Climate Change Initiative and Copernicus Earth observation program). The building blocks in this procedure will be lessons learned from previous Cal/Val studies (including CALIPSO, Cloudsat, GPM, and Aeolus), as well as the well-established QC/QA procedures adopted by the related European Research Infrastructures and metrological institutes (e.g., ACTRIS i.e. Aerosol Cloud and Trace Gasses Research Infrastructure, WMO-WRDC i.e. World Meteorological Organization - The World Radiation Data Centre). The approach will evolve from a review of the current literature, and will be consolidated in consultation with the community at workshops and via the EarthCARE Validation portal.
The presentation will address the development status of the protocols, and explain how the broader community can participate in their formulation. Contributions from the cloud and aerosol communities are expected to gradually broaden the coverage of the validation protocols. While initially focusing on EarthCARE, the best validation practices could be extended to other current and future missions (e.g., ESA Aeolus and its follow on mission, NASA EOS/AOS i.e. Earth System Observatory / Atmosphere Observing System, and WIVERN i.e. WInd VElocity Radar Nephoscope).

Clouds and aerosols play an essential role in the Earth's radiative balance and therefore condition its temperature and possibly its evolution. Knowledge of their life cycle is therefore essential to understand the Earth's climate but also to predict meteorological conditions. The EarthCARE mission, currently under development for launch in 2023, was decided in this direction to provide solutions to these questions. To do this, it will probe the Earth's atmosphere by measuring the profiles of clouds and aerosols but also radiation thanks to its set of on-board instruments including radar and lidar. The association of these two instruments is not accidental. Indeed, the synergy of radar and lidar measurements collocated over an area or a transect of the atmosphere is a powerful tool for removing any ambiguity about the atmospheric targets present. The AC-TC (ATLID CPR – Target Classification) product is then created in this goal for EarthCare. It is a synergistic product that combines observations from the Cloud Profiling Doppler Radar (CPR) and the high spectral resolution Atmospheric Lidar (ATLID) on board EarthCARE satellite (ESA-JAXA). This product relies on the complementary nature of radar and lidar measurements to properly define targets (hydrometeors and aerosols) present when probing the atmosphere. Each instrument is sensitive to different parts of particle size regimes, with ATLID probing the smaller particles (i.e. aerosols and cloud particles) and CPR more sensitive to the larger particles (i.e. ice cloud particles and precipitation) providing independent information (microwave or optical) in the region of overlap. The combination of their signals makes it possible to better classify the different atmospheric targets in comparison to the single instruments. Therefore, the cloud phase, precipitation and aerosol type within the column sampled by the two instruments can be identified. This product is a crucial step for the subsequent synergistic retrieval of cloud, aerosol and precipitation properties. Further, it can also be used on its own for statistical studies of atmospheric conditions via e.g. the statistical analysis of the cloud, aerosol and precipitation occurrence. The AC-TC product capitalizes on the enormous success of CloudSat/CALIPSO satellites on the A-Train constellation and their synergistic derivatives while providing a richer target classification due to the EarthCARE instruments.

The great benefit of EarthCARE for the global observation of the atmosphere lies in its synergistic approach of combining four instruments on one single platform. The two active instruments ATLID (atmospheric lidar) and CPR (cloud profiling radar) deliver vertical profiles of aerosol and cloud properties. The two passive instruments BBR (broad band radiometer) and MSI (multispectral imager) extend the information by adding observations of the total and shortwave radiation at top of atmosphere (BBR) and spectral radiances at the across track swath (MSI).
The systematic combination of active and passive remote sensing on a single platform is new and offers us great opportunities for synergistic retrieval approaches. Here, we will focus on the synergy of the vertical profiles measured with ATLID (‘curtain’) and the horizontal information added by the MSI (‘carpet’) to provide a more complete picture of the observed scene. For this purpose, the synergistic ATLID-MSI columnar descriptor AM-COL was developed in the EarthCARE processing chain. The MSI input is provided by the MSI cloud (M-CLD) and MSI aerosol (M-AOT) processor, the ATLID input is calculated by the ATLID layer processor (A-LAY). Cloud and aerosol information derived on the track are combined from both instruments and the additional ATLID information is transferred to the swath using the MSI observations. Two main results will be described in the following paragraphs, the cloud top height and the Ångström exponent.
The difference of the cloud top height measured with ATLID and retrieved from MSI is calculated along track. The obtained differences are transferred to the swath searching for similar nearby MSI pixels. Five homogeneity criteria are used: same cloud type, same cloud phase and surface type, the reflectivity at 0.67 µm and the brightness temperature at 10.8 µm. At nighttime, only a reduced set of criteria can be used (no cloud type and no reflectivity differences are available). Multilayer cloud scenarios have to be treated with special care and are investigated separately. Especially, thin cirrus clouds above liquid-containing clouds are hardly detectable with MSI.
The three simulated test scenes developed for EarthCARE are intensively used to test the algorithm performance. In case of the mixed-phase clouds and some thick cirrus clouds present in the so-called ‘Halifax’ scene, the difference between ATLID and MSI is found to be smaller than 1000 m. When looking at multilayer clouds or large convective systems this difference increases. For homogeneous cloud coverage, the transfer of the cloud top height difference to the swath can be easily applied. The test scenes offer the possibility to check the transfer to the swath for the more complicated multilayer scenes as well. The comparison to the model truth lets us estimate the performance of the synergistic product and provides an estimation for the detection limits when it comes to real data.
The aerosol optical properties are obtained at 670 nm and 865 nm (ocean only) by MSI and at 355 nm by ATLID. The ATLID-MSI synergy enables us to calculate the Ångström exponent (355/670 and 355/865) along track adding spectral information to the single wavelength lidar ATLID. The Ångström exponent adds additional information to the aerosol typing. Along the nadir track we can combine the vertical resolved aerosol classification from ATLID with the aerosol typing included in the MSI retrieval. Knowing the aerosol type along track seen by both ATLID and MSI enables us to transfer the aerosol information to the swath using the MSI measurements. For this purpose, an explicit aerosol test scene was developed additionally to the three standard EarthCARE test scenes. It could be shown that the MSI-based aerosol typing agrees with the columnar aerosol classification probabilities derived from ATLID for this scene.

The Earth Clouds, Aerosols and Radiation Explorer (EarthCARE) has the scientific goal to achieve agreement of +- 10 W/m² for average SW/LW fluxes simulated using radiative transfer models acting on the retrieved profiles of cloud and aerosol properties and values inferred from collocated measurements made by the broadband radiometer (BBR).

The fluxes are estimated from BBR measurements at a single sun-observer geometry of the satellite using angular distribution models (ADMs). ADMs for SW radiances are created for different scene types and constructed from Clouds and the Earth’s Radiant Energy System (CERES) data using a feed-forward back-propagation artificial neural network (ANN) technique (Domenech et al., 2011) .
To further improve the solar flux estimates, a new method has been developed to possibly supplement the ANN technique (Tornow et al., 2020). The semi-physical log-linear approach incorporates cloud effective radius (Reff) and cloud topped water vapor as additional parameters which can significantly influence the TOA solar flux through changes in scattering direction and absorption respectively. A comparison with the state-of-the-art solar flux retrievals obtained from CERES and GERB instruments showed significant flux differences for cloudy scenes over ocean, which has been attributed to extremes in Reff and cloud topped water vapor (Tornow et al., 2021).

In the study presented here, the new method is evaluated and compared with the ANN technique. Since EarthCare is not yet in orbit, simulated EarthCare frames (1/8 of the orbit) are used. The frames were created by the EarthCare team using the Global Environmental Multiscale (GEM) model from Environment Canada and Climate Change and ESA instrument models.
Situations with large differences are analysed and interpreted in more detail. Further it is discussed in which situations the ANN technique could be complemented by the new method.

The upcoming EarthCARE mission will deliver horizontal and vertical aerosol information from one single platform. While ATLID (atmospheric lidar) will be responsible for vertically resolved aerosol properties, horizontal, columnar information about aerosol will be provided by MSI (multi-spectral imager) measurements. For the latter, the L2 aerosol processor M-AOT has been developed. It will operationally estimate aerosol optical thickness over ocean at 670 and 865 nm and, where possible, over land at 670 nm.

Measurements of the four available MSI bands in the visible to shortwave infra-red (670 nm, 865 nm 1650 nm and 2200 nm) are used within the underlying algorithm that consists of a separate land and ocean retrieval part. The ocean surface is parameterized following Cox and Munk (1954) and the land surface albedo is empirically parameterized relying on information about the vegetation type and the albedo at 2200 nm. Both algorithm parts are using an optimal estimation framework, whose forward operator relies on pre-calculated look-up tables that have been generated using radiative transfer code MOMO [Hollstein and Fischer, 2012] and are using the EarthCARE Hybrid End-To-End Aerosol Classification (HETEAC) model [Wandinger et al. 2016] to ensure consistency between ATLID and MSI based aerosol products.

Here, the underlying algorithm itself and product examples based on EarthCARE simulator test scenes will be presented together with algorithm verification and already known limitations of the area of application based on retrieval testing with MODIS input data.


Cox, C. and Munk, W.: Measurements of the roughness of the sea surface from photographs of the sun's glitter. J.Optical Soc. Amer. 44, Pages 838-850, 1954.
Hollstein, A. and Fischer,J.: Radiative transfer solutions for coupled atmosphere ocean systems using the matrix operator technique. Journal of Quantitative Spectroscopy and Radiative Transfer Volume 113, Issue 7, May 2012, Pages 536–548, 2012.
Wandinger, U., Baars, H., Engelmann, R., Hünerbein, A., Horn, S., Kanitz, T., Donovan, D., van Zadelhoff, G.J., Daou, D., Fischer J., von Bismarck, J., Filipitsch, F., Docter, N., Eisinger, M., Lajas, D. and Wehr, T.: HETEAC: The Aerosol Classification Model for EarthCARE. EPJ Web of Conferences, 119, 2016

Validation and calibration techniques for advanced space borne Earth observation instruments usually relies on ground based reference instruments to provide reference measurements able to assess the performance of the corresponding space instrument. The newly developed A-lidar instrument is one of the reference lidar instrument designed to meet all recommended requirements of the European Research Infrastructure for Short-lived Atmospheric Constituents - ACTRIS (actris.eu). The instrument aims to provide continuous data and will provide data to a wide range of users in order to facilitate high-quality Earth climate research.
The Alpha-lidar is designed to provide daytime backscatter, one daytime extinction, nighttime extinction and three depolarization products (3β, 1α-daytime + 1hsrl, 6α-nighttime, 3δ, 1 water vapour). To achieve these specifications, the instrument makes use of the rotational and vibrational Raman lines at 355, 532 and 1064nm. The instrument is designed to achieve full overlap around 200m for the primary lidar products like the raw and backscatter profiles and can go to lower altitudes for products where signal ratios are used (like the depolarization products).
The Alpha-lidar is split in an operational and an experimental part. The operational part is made up from three lasers and three telescopes, each emitter/receiver pair focusing on a different atmospheric property. The first receiver covers the elastic and Raman channels, the second is dedicated to the 532 and 355nm depolarization channels and the third receiver is dedicated to the 1064nm depolarization channels. In addition to the main lidar units, the instrument is also equipped with an experimental HSRL unit based on the Iodine filtering technique for 532nm. The entire instrument is enclosed in a custom container designed to accommodate the instrument for continuous operation in all weather conditions – see Figure 1.
The data retrieved with the instrument indicates good operation for both daytime and night-time setup. Once all quality assurance tests will be finalized, the instrument will be set for operational use and will be included in the operational programme of the Actris-RI as one of the reference instruments of the CARS central facility (Centre for Aerosol Remote Sensing).
The instrument could be one of the tools used in the EarthCARE and other similar mission in the validation programme. During the conference, several lidar derived products and associated errors, highlighting different atmospheric features will be presented. Product examples retrieved using the ACTRIS Single Calculus Chain are presented in Fig.2.

Acknowledgements:
The work performed for this study was funded by the Ministry of Research and Innovation through the Romanian National Core Program Contract No.18N/2019 and by the European Regional Development Fund through Competitiveness Operational Programme 2014–2020, POC-A.1-A.1.1.1- F- 2015, project Research Centre for environment and Earth Observation CEO-Terra.The research leading to these results has received funding from the European Union H2020, ACTRIS IMP grant no. 871115.

The Earth Clouds, Aerosols and Radiation Explorer (EarthCARE) mission will carry a depolarization-sensitive high-spectral-resolution lidar as well as a Doppler radar for global measurements of aerosol and cloud properties. These observations will be used in radiative transfer simulations to pursue the main objective of the mission: the radiative closure of the Earth’s radiation budget at top-of-the-atmosphere (TOA) using complementary on-board passive remote sensing observations for comparison. To achieve best possible agreements between the derived radiative fluxes from active remote sensing and passive measurements, the distributions of radiatively active constituents of the atmosphere have to be known. Especially the vertical distribution of water vapor should be precisely characterized, as it is spatially and temporally extremely variable. However, with water vapor profiles not being directly measured by EarthCARE, radiative transfer models have to rely on modeled vertical atmospheric water vapor distributions and standard atmospheric profiles.
During two airborne research campaigns over the western Atlantic Ocean, we conducted lidar measurements aboard the German HALO (High Altitude and Long Range) research aircraft above transported Saharan dust layers. All measurements indicated enhanced concentrations of water vapor inside the dust layers compared to the surrounding free atmosphere. We found that the embedded water vapor in the dust layers has a great effect on the vertical heating rate profiles as well as on TOA radiation. Hence, with the main goal of EarthCARE being the closure of the Earth’s radiation budget at TOA, particular attention has to be payed to a correct parametrization of the vertical water vapor profile and its possible radiative effects.
In our presentation, we will present the derived radiative effects of long-range-transported Saharan dust layers from EarthCARE-like remote-sensing with HALO during both boreal winter and summer. We will highlight the contribution of enhanced concentrations of water vapor in the dust layers to calculated TOA radiative effects as well as heating rates. Additionally, we compare our results to radiative transfer calculations where standard distributions of water vapor are used.

The EarthCARE satellite mission targets an improved understanding of the influence of clouds and aerosols on the global radiation budget. Toward this goal, a target accuracy of +/- 10 W/m2 has been defined as threshold for closure between observed top-of-atmosphere fluxes and 3D radiative transfer simulations on spatial domains with an area of 10x10 km2. For our understanding of climate processes and other applications, closure of surface radiative fluxes is also of critical importance, but is not currently covered by the EarthCARE mission concept. Radiative closure is however much more difficult to assess experimentally at the surface than at the top-of-atmosphere in particular due to the limited spatial representativeness of ground-based measurements for larger domains, if instantaneous fluxes are considered. A common approach is to average observations for longer time periods or a large number of similar situations to reduce this sampling uncertainty, but this approach is also susceptible to error cancellation. An alternative is the deployment of a dense network of radiation sensors to better sample the average radiation fluxes across a region of interest. A key advantage is the possibility to investigate deviations and assess closure on a case-by-case basis. Using observations from several past field campaigns with a low-cost pyranometer network, the feasibility of such a closure experiment for surface radiative fluxes based on EarthCARE products and processors is assessed. A method based on optimum averaging/ spatio-temporal Kriging is introduced to determine the sampling accuracy of a sensor network for domain-average instantaneous fluxes. For several typical cloud situations, the number of stations required to reach different target accuracies for the average flux across the EarthCARE closure domain size is determined. Based on these findings, potential instrumental configurations for such an experiment are described.

Aboveground forest biomass (AGB) accounts for between 70% to 90% of total forest biomass estimates, which are the central basis for carbon inventories. Estimation of forest aboveground biomass (AGB) is critical for regional forestry and sustainable forest management. Remote sensing (RS) data and methods offer opportunities of AGB broad-scale assessments providing data over large areas at a fraction of the cost with access to inaccessible places. Optical RS provides good alternative to biomass estimation through field sampling due to its global coverage, repetitiveness and cost-effectiveness. Radar RS has gained prominence for AGB estimation in recent years due to its cloud penetration ability as well as detailed vegetation structural information.
In this study, the potential of C-band SAR data from Sentinel-1, L-band SAR data from ALOS PALSAR, multispectral data from Sentinel-2 instruments and machine learning algorithms were evaluated for the estimation of AGB in a mountainous mixed forest in the Eastern part of the Czech Republic. The response variable was AGB (Mg/ha) estimated from normalized digital surface model nDSM (Forest Management Institute, http://www.uhul.cz) and field measurements (R2=0.84, nRMSE = 10%). The following cases of predictors were considered for AGB modelling: (1) Sentinel-1, Sentinel-2 and ALOS PALSAR, (2) Sentinel-1 and Sentinel-2, (3) ALOS PALSAR and Sentinel-2. SAR data were used with VV and VH polarizations. Normalized difference vegetation index NDVI, tasselled cap transformation TC (greenness, brightness and wetness) and disturbance index DI were calculated from multispectral Sentinel-2 data and together with single spectral bands were used as predictors. The modeling was performed with several machine-learning algorithms including, neural network, adaptive boosting and random decision forest. The AGB models were developed for coniferous, deciduous and mixed types of forest. AGB estimates for deciduous forest stands generally showed a weaker predictive capacity for all models, than AGB estimates for coniferous. The models with Sentinel-1 and Sentinel-2 predictors (case 2) had the weaker estimates comparing with models using ALOS PALSAR predictors (cases 1 and 3). The best model performance was achieved with the random decision forest algorithm and predictors derived from three sources of satellite data, Sentinel-1, Sentinel-2 and ALOS PALSAR. The proposed methodology can be applicable for Central European forest AGB mapping in large areas using the satellite optical and radar data.

Keywords: machine learning, ALOS PALSAR, Sentinel, forest productivity.

Acknowledgment: The study was supported by the Ministry of Agriculture of the Czech Republic, grant number QK1910150.

References:
Cairns et al. 2018. Root biomass allocation in the world’s upland forests. Oecologia 1997, 111, 1–11.
Chen et al. 2018. Estimation of forest above-ground biomass by geographically weighted regression and machine learning with Sentinel imagery. Forests, 9.
Pedregosa et al. 2011. Scikit-learn: Machine Learning in Python. Journal of Machine Learning Research, 12.
Perko et al. 2011. Forest assessment using high resolution SAR data in X-band. Remote Sens., 3, 792-815.

For the past decades, wildfires have been increasing in frequency and in severity worldwide. These fires are source of substantial quantities of CO2 released in the atmosphere. They can also lead to the destruction of natural ecosystems and biodiversity. Fires are triggered by various factors that depend on the climate regime and on the vegetation type. Despite the large number of studies conducted on wildfires, post-fire vegetation recovery is still to be better understood, and depends highly on the vegetation type.
In this study, we present pre and post fire climate and vegetation anomalies at global scale, from several remotely sensed observations, such as air temperature (MODIS), precipitation (PERSIANN-CDR), soil moisture (SMOS), and terrestrial water storage (GRACE). Four remotely sensed variables related to vegetation are used and compared, from optical (the enhanced vegetation index (EVI) from MODIS) to microwave wavelengths opacities ranging from 2 to 20 cm: X-band, C-band, and L-band vegetation optical depth (X-VOD, C-VOD, and L-VOD), obtained with AMSR-2 and SMOS satellites. Fires are detected with the MODIS Active Fire product (MOD14A1_M). All datasets are resampled to SMOS grid (~ 25 km) and at a monthly timescale, for the time period June 2010 – December 2020.
We focus our analysis on five particular biomes : grasslands, tropical savannas, needleleaf forests, sparse broadleaf forests, and dense broadleaf forests. Anomalies of all data are computed over the major fires of the ten-year period, at global scale, then time series are readjusted on the fire date and averaged by biome.
We observe a severe drought before the majority of the fire events, and in particular over forests, which generally maintain a steady humidity all year. Pre-fire temperature anomalies are particularly significant in boreal needleleaf forests. In contrast, over savannas and grasslands, the pre-fire drought is slight while an increase in the biomass volume (e. g., available fuel) is supposed to expedite fires. As expected, C- and X-bands are more affected by sparse vegetation fires, as these frequencies are sensitive to the smaller branches and leaves; whereas L-band is particularly impacted over dense broadleaf forest fires, as it is a measurement of coarse woody elements (trunks and stems). For all biomes, the optical-based index (EVI) decreases significantly after fire but recovers quickly, as it observes only herbage and green canopy foliage. The contrasted recovery duration between L-VOD and the other variables over dense forests shows that fires affect coarse woody elements in the long term, while stems and leaves resprout faster. Our study shows the potential of SMOS L-VOD to monitor fire-affected areas as well as post-fire recovery, especially over densely vegetated areas. This study is also the first one to compare multi-frequencies VODs and to observe the impact of fire in L-VOD signal.

Figure - EVI, X-, C-, L-VOD, precipitation, SM, TWS, and temperature anomalies time series, shifted on the fire date, for (a) 520 points in the grassland biome; (b) 232 points in the savanna biome; (c) 701 points in the needleleaf forest biome; (d) 69 points in the sparse broadleaf forest biome; and (e) 48 points in the dense broadleaf forest biome. The missing values are mainly due to snow filtering.

The ability to capture 3D point clouds from LiDAR sensors and the advancement in algorithms has enabled the explicit analysis of vegetation architecture, branching characteristics and crown structure for the accurate estimation of Above Ground Biomass (AGB). The ability to have geometrically accurate 3D volume of vegetation reduce the uncertainty in AGB estimation without destructive sampling through the application of volume reconstructions algorithms on high-resolution point clouds from Terrestrial Laser Scanning (TLS). These methods, however, have been developed and tested on temperate and boreal vegetation with very little emphasis on the savanna vegetation. Here, we test the reconstruction algorithms for the estimation of AGB in a savanna ecosystem characterised by a dense shrub understory and irregular multi-stemmed trees. Leaf off multi scan TLS point clouds were acquired during the dry season in 2015 around the Skukuza flux tower in Kruger National Park, South Africa. From the multi scan TLS point clouds, we extracted individual tree and shrub point clouds. Tree Quantitative Structure Models (TreeQSMs) were used to reconstruct tree woody volume whilst voxel approaches were used to reconstruct shrub volume. The AGB was estimated using the derived woody volume and wood specific gravity. To validate our method, we compared the TLS derived AGB with allometric equations. TreeQSMs predicted AGB with a high concordance correlation coefficient (CCC) compared to the allometry reference, although tree crown biomass was overestimated especially for the large trees. The biomass of the shrub understory was described with reasonable accuracy using the voxel approach. These findings indicate that the application of 3D reconstruction algorithms improve the estimation of savanna vegetation AGB as compared to allometry references and combined tree and shrub woody biomass estimates of the savanna allow for calibration and validation for accurate monitoring and mapping at large spatial scales.

Tropical dry forests harbor major carbon stocks but are rapidly disappearing due to agricultural expansion and forest degradation. Yet, robustly mapping carbon stocks in tropical dry forests remains challenging due to the structural complexity of these systems on one hand, and the lack of ground data on the other. Here we combine data from optical (MODIS) and radar (Sentinel 1) time series, along with Lidar-based (GEDI) canopy height information, in a Gradient Boosting Regression framework to map aboveground biomass (AGB) in tropical dry forests. We apply this approach across the entire dry Chaco ecoregion (800,000 km2) for the year 2019, using an extensive ground dataset of forest inventory plots for training and independent validation. We then compare our AGB models to structural vegetation parameter such as percent tree and shrub cover, as well as Level-2 data from GEDI. Our best AGB model considered MODIS and Sentinel 1 data, whereas the additional use of GEDI-based canopy height data did not contribute substantially to model performance. The resulting map, the first high-resolution AGB map covering the entire ecoregion, revealed that there are still 4.65 Gt (+/- 0.9 Gt) of AGB in the remaining natural woody vegetation of the Chaco. Nearly three quarter of the remaining AGB in natural vegetation is located outside protected areas, and nearly half of the remaining AGB occurs on land utilized by traditional communities, suggesting considerable co-benefits between protecting traditional livelihoods and carbon stocks. Our models also had a much higher level of agreement with independent ground-data than global AGB products, which translates into a huge, up to 14-fold, underestimation of AGB in the Chaco by global maps in comparison to our regional product. Our map represents the most accurate and fine-scale map for this global deforestation hotspot and reveals substantial risk of continued high carbon emissions should agricultural expansion progress. In addition, by combining our AGB map with structural vegetation parameters we provide for the first time for tropical dry forests an understanding of carbon stocks in relation to the vegetation structure in these ecoregions. More broadly, our analyses reveal the considerable potential of combining time series of optical and radar data for a more reliable mapping of above-ground biomass in tropical dry forests and savannas.

Forests play a critical role in the global carbon cycle. However, estimates of forests carbon storage still have large uncertainties, especially in tropical forests. In addition, the distribution of above-ground biomass (AGB) at certain heights in forests (vertical AGB distribution) is completely underexplored at large scales with remote sensing. Synthetic aperture radar (SAR) and light detection and ranging (lidar) are common remote sensing tools used to estimate AGB. SAR has a large coverage imaging capability, and lidar can achieve high accuracy for measuring forest structure. The tomographic SAR mode (TomoSAR) of ESA’s upcoming P-band SAR satellite BIOMASS together with NASA’s Global Ecosystem Dynamics Investigation (GEDI) spaceborne lidar system will provide an unprecedented opportunity to estimate the vertical distribution of AGB at a regional or global scale. Our objective in this study was to develop and evaluate an approach to estimate the vertical distribution of AGB by combining observations from GEDI and a TomoSAR system (DLR’s airborne F-SAR) for the forest sites in Lopé and Mondah, Gabon, Africa.
According to the ESA WorldCover 10 m 2020 product, the research area in Lopé is covered by 79% trees and 19% grassland. The research area in Mondah is covered by 76% trees, 5% grassland, 15% permanent water bodies and 2% built-up. We used P-band TomoSAR data from the F-SAR system acquired during the ESA AfriSAR 2016 campaign, and GEDI level 2A (ground elevation, canopy top height, relative height metrics) and level 4A (footprint level above ground biomass) products. GEDI data were filtered based on the available quality flags and sensitivity metrics. There were 1,446 and 182 filtered GEDI footprints at 25 m resolution in Lopé and Mondah, respectively. Airborne lidar data from NASA Land, Vegetation, and Ice Sensor (LVIS) was used as reference.
Firstly, we applied the Capon method to reconstruct the reflectivity profiles from 10 tracks HV polarised P-band SAR images. We normalised the tomographic intensities into [0, 1] and used 0.1 as minimum threshold to cut profiles. The lowest peak of each profile was regarded as ground (relative height, RH0). The position above the highest peak and with intensity equal to 0.1 was selected as the RH100, considering the penetration capability of P-band microwave. The relative height (RH) retrieved from GEDI, TomoSAR and LVIS were compared at 25 m and 200 m spatial resolution, representing the resolution of LVIS and GEDI height products, and the resolution of future BIOMASS height products, respectively. Instead of using common height-AGB allometry relationships or power law models based on TomoSAR intensity at certain height level (e.g., 30 m), we attempted here to estimate total AGB from the TomoSAR profile directly. This approach also enables us to quantify the contribution of different TomoSAR height levels to the estimation of total AGB. Therefore, with GEDI AGB as response, random forest regression was then applied to estimate total AGB from TomoSAR profiles at 50 m (resolution of LVIS AGB product) and 200 m resolution (resolution of BIOMASS AGB product). The input features are TomoSAR intensities from 0 to 60 m in 5 m steps. Theses profiles were subset to start from R0 and the intensities above RH100 were set to zero for ensuring the fixed length (i.e., 13) of predictors. The samples were split into training set (80%) and testing set (20%). A five-fold cross-validation was carried out to test model’s transferability and the model with highest coefficient of determination (R²) from five cross-validation models was selected as the final model. In order to estimate the vertical distribution of AGB, we combined in-situ measurements and data from the Biomass And Allometry Database (BAAD) to describe the vertical AGB distribution of individual trees. Therefore, the crown was modelled as a sphere and the stem was modelled as a cone. These individual AGB profiles were then summed up to get the AGB profile at the plot scale. An optimal extinction factor for the P-band microwaves was estimated based on the root-mean-square-error (RMSE) between TomoSAR profiles and normalised field AGB profiles at grid level. Considering the discrepancy between TomoSAR RH100 and in-situ measured (or modelled) forest height, we subset the TomoSAR profiles corresponding to field plots using in-situ forest height rather TomoSAR RH100. By combining the estimate of total AGB with the optimised extinction factor and the TomoSAR profiles, we derived the vertically distributed AGB of the whole research area at 50 m resolution.
Our results show that the RH metrics from GEDI, TomoSAR and LVIS match well in the two study areas. For the cross-validation of the random forest model, models for Lopé (R² = 0.77) and Mondah (R² = 0.81) have similar performance at 50 m resolution, while model for Lopé (R² = 0.86) at 200 m performs better than that for Mondah (R² = 0.77). In both study sites, the R2 between predictions and reference data at 200 m resolution is around 0.2 higher than the R2 at 50 m resolution when these models are extended to the whole research area. The feature importance of random forest model in Lopé and Mondah show that the tomographic intensity between 20 m and 40 m contribute most to the total AGB. From the perspective of normalised root-mean-square (NRMSE), the forest height estimated from TomoSAR satisfies the requirement of the BIOMASS mission (BIOMASS: 30%, Lopé: 13%, Mondah: 11%), while the AGB from TomoSAR does not (BIOMASS: 20%, Lopé: 26%, Mondah: 28%). With an optimal extinction factor, the mean R² between reconstructed TomoSAR AGB profiles with their counterparts derived from field observations is 0.7. In summary, our results demonstrate the potential of combining spaceborne lidar measurements with future spaceborne TomoSAR measurements to get a more detailed insight in the vertical distribution of biomass in tropical forests and understand performance limitations of prospective BIOMASS products.

Fire risk assessment in forest stands relies on detailed information about the availability and spatial distribution of fuels. In particular surface fuels such as litter, downed wood, herbs, shrubs and young trees determine fire behaviour in temperate forests and constitute the primary source of smoke emissions. Remote sensing has been suggested as a potentially valuable tool to estimate the spatial distribution of fuels across large areas. However, accurately estimating surface fuel loadings in space across various fuel components using airborne or spaceborne sensors is complicated by obstruction from the forest canopy. In addition, mapping efforts have largely focused on simplified representations of fuel situations for specific modelling purposes, such as classifications into fuel types or fuel models, rather than estimating fuel loadings. In this work, we test whether the fusion of high-resolution LiDAR data (> 60 points/m²) with moderate- to high-resolution satellite imagery from the Sentinel-2 mission (10-20 m) allows to predict loadings of all surface fuel components using machine learning techniques. Our analysis is based on a field inventory of surface fuels in a mixed temperate forest with two dominating deciduous tree species (Fagus sylvatica, Quercus petraea) and two dominating coniferous species (Pinus sylvestris, Pseudotsuga menziesii). We produce fine-scale maps of surface fuel loadings that can form the basis for fuel management strategies as well as for calculations of fire behaviour characteristics and fire effects. Furthermore, we test how spatial variability in surface fuel loadings is captured when broader categories such as fuel types are used as mapping units. We investigate possible relationships of overstory tree species and cover with surface fuel loadings to reach more general conclusions about predictors for surface fuel loadings in temperate forests of central Europe. Our study contributes to a better understanding of fuel-related fire risk in temperate forests, which can help in developing appropriate forest management decisions and fire-fighting strategies.

Forests are essential in maintaining healthy ecosystem interaction on Earth. Forests cover around 30% of the world’s land area (FAO 2020), are estimated to contain over 500 Pg of above ground live biomass (Santoro et al. 2021) and represent a net sink of −7.6 ± 49 GtCO2e yr−1 (Harris et al. 2021). This makes them a crucial asset in the fight against climate change. Reliable information on forest biomass and carbon fluxes are needed to meet the reporting requirements of national and international policies and commitments like the Paris Agreement on Climate Change and the United Nations’ Sustainable Development Goals (Herold et al. 2019).

The Forest Carbon Monitoring project (https://www.forestcarbonplatform.org/) for the European Space Agency is developing Earth Observation (EO) based user-centric approaches for forest carbon monitoring. Different stakeholders have a common challenge to monitor forest biomass, but specific requirements vary between users. Policy-makers need information to make better decisions; public organizations need information for national and international level reporting. Companies require means to respond to increasing monitoring requirements, and tools for carbon trading. To support forestry stakeholders in these requirements, the project aims to develop a prototype of a monitoring platform which offers:

• A selection of statistically robust monitoring approaches designed for accurate forest biomass and carbon monitoring for varying large and small area requirements.
• Cloud processing capabilities, unleashing the potential of the increased volumes of high-resolution satellite data and other large datasets for forest biomass and carbon monitoring

In this presentation, we will give an overview of the project’s status, first results and further development. We will specifically highlight the research efforts to be undertaken in this project to improve usability of Earth Observation (EO) in meeting the varying user needs in forest biomass and carbon monitoring. The project started with an extensive review of policy needs and users’ technical, methodological and data requirements. Project user partners were interviewed for detailed requirements. This information was reflected against the current state-of-the-art of EO based forest carbon monitoring methods to identify the potential and limitations of EO based forest biomass monitoring. During the first year of the project, different approaches for data processing, biomass estimation and uncertainty assessment have been tested and evaluated. During the second year of the project, three different types of demonstrations will be conducted and validated:

• Local level demo designed to meet private company and other small area requirements.

• Provincial to national level demo aimed primarily at administrative agencies, often using National Forest Inventory (NFI) based approaches.

• Continental level demo, aiming to meet the needs of international organizations and other communities requiring continental level information.

The underlying policy and user requirements analysis including user interviews highlighted the variety of requirements that forestry stakeholders have towards forest monitoring in general and with a focus on biomass carbon. The needs could be coarsely grouped according to the three different types of demonstrations. Particularly the private companies with smaller interest areas need basic forest structural variables (e.g. basal area, diameter, height, volume) as much, or even more, than forest biomass and carbon data. These basic forest variables support their forest management decisions, but also allow biomass or carbon flux estimation when required. Public and international users, on the other hand, are more specific in their requirements regarding the variables or interest, as they are defined by the policy reporting requirements. For the national level organizations, existing approaches are heavily based on NFI field data, and any supporting EO based approaches need to be able to complement the existing monitoring systems in a productive manner. All users raised the importance of reliable and accurate monitoring and reporting of changes in forest biomass.

Due to the large amount of research conducted on the increased volume and variety of high spatial resolution EO data (see e.g. Miettinen et al. 2021), combined with processing capabilities enabled by cloud processing environments, the scientific readiness for EO based forest biomass monitoring is rising fast to the level required meet the user requirements. However, not all of the approaches are ready for operational use and should be further developed. Particular attention needs to be given to the fact that in operational circumstances the available datasets and monitoring conditions are rarely optimal, affecting the quality and consistency of the outputs. Key research issues that need more investigation to properly respond to the user requirements include:

• In an operational system responding to user needs, robust and transparent uncertainty assessment approaches and validation procedures are crucial. Reference data availability is rarely optimal in operational setting, requiring development of several uncertainty monitoring approaches validation procedures to be applied according to available datasets.

• In direct growing stock volume and biomass estimation, further development is needed on utilization of multi-temporal and multi-sensor datasets, combined with improved model calibrations. Approaches such as developed by Santoro et al. (2021) have proven useful for global level analyses, improved pixel level accuracies would enable derivation of reliable results for smaller interest areas and comparison between two time steps of mapping.

• For basic forest structural variable estimation, the availability and suitability of field reference measurements is a crucial issue and better integration with NFI data should be sought. Further improvements are also pursued e.g. from combined use of optical and radar datasets, as well as utilization of variable-specific estimation methods.

• A key feature of the platform to be developed in this project is integration of ecosystem simulation models into the system. The calibration of these models for different tree species and site conditions is still a significant knowledge gap even for European and particularly for global application. By means of data assimilation, the utilization of a modelling framework allows also to integrate multiple data sources for forest monitoring, enabling set-up of a continuously updating monitoring system. This is a major area of development in the long run for forest biomass and carbon monitoring.

The main results achieved in each of the research areas listed above will be reported in the presentation. Overall, the required knowledge on what needs to be done to set up the planned Forest Carbon Monitoring platform exists. However, the research currently conducted in the above listed topics aims to improve the reliability, usability and integration of EO based approaches to such a level that would enable wider onboarding of EO based approaches for forest biomass and carbon monitoring by forestry stakeholders. Although this project focuses on biomass and carbon monitoring, a forest monitoring platform should ultimately have a broader focus than carbon and cover effects of climate change, biodiversity, health, damages, invasive alien species, forest management, and the biomass use.


References:

FAO (2020) The State of the World's Forests 2020: In brief – Forests, biodiversity and people. Rome: FAO & UNEP. doi: 10.4060/ca8985en. ISBN 978-92-5-132707-4.

Harris, N.L., Gibbs, D.A., Baccini, A. et al. (202) Global maps of twenty-first century forest carbon fluxes. Nature Climate Change 11, 234-240. doi: 10.1038/s41558-020-00976-6.

Herold, M., Carter, S., Avitabile, V. et al. (2019) The Role and Need for Space‑Based Forest Biomass‑Related Measurements in Environmental Management and Policy. Surveys in Geophysics 40: 757–778. doi: 10.1007/s10712-019-09510-6.

Miettinen, J., Rauste, Y., Gomez, S., et al. (2021) Compendium of Research and Development Needs for Implementation of European Sustainable Forest Management Copernicus Capacity; Version 2. Available at: https://www.reddcopernicus.info/wp-content/uploads/2021/06/REDDCopernicus_RD_Needs_SFM_V2.pdf

Santoro, M., Cartus, O., Carvalhais, N. et al. (2021) The global forest above-ground biomass pool for 2010 estimated from high-resolution satellite observations. Earth System Science Data 13: 3927–3950. doi: 10.5194/essd-13-3927-2021

Modern agriculture should combine the needs of productivity with those of environmental, economic and social sustainability, in an uncertain climate context due to the effects of climate change. Information useful for implementing advanced and integrated monitoring and forecasting systems to promptly identify the risks and the impacts of calamities and crop practices on agricultural environments are essential. Satellite Earth observation data revealed to be optimal for the aforementioned tasks because they can cover wide areas with different spatial resolutions and frequent revisit time, allowing the collection of historical series for long-term analysis, and they can be punctual thanks to the continuous acquisition of Copernicus constellations. Finally, from an economic point of view they are becoming more convenient thanks to the provision of free satellite data and dedicated software for their processing and display.
Agricultural ecosystems are characterized by strong variations within relatively short time intervals. Depending on the observation period the agricultural scenario can present itself in a totally different way, due to the difference in biomass and phenological cycle, that can be driven by cultivar and agricultural working, as well as weather conditions. These dynamics are challenging for crop monitoring and the knowledge of vegetation status can deliver crucial information that can be used to improve the classifiers performance.
In order to consider these aforementioned changes in agricultural vegetation and soil status, a multitemporal approach based on the study of time series of SAR indices can be successful. Time series of satellite images offer the opportunity to retrieve dynamic properties of target surfaces by investigating their spectral properties combined with temporal information on their changes.
This research work was carried out using SAR images from Sentinel-1 (at VV and VH polarizations) and COSMO-SkyMed (at HH polarization for Himage and VV+HH polarizations for PingPong) satellite sensors, which have been collected for a few years over an agricultural area in central Italy. The sensitivity of backscattering and related polarization indices at both C and X bands was investigated and assessed in several experiments. Both frequencies revealed to be sensitive to crop growth although with different behaviors according to crop type, the backscatter being influenced by the two phenomena of absorption and scattering caused by the dimensions of leaves and stems. In particular, crops characterized by large leaves and thick stems cause an increasing of backscattering as the plants grow and the biomass increases; whereas crops characterized by narrow leaves and thin and dense stems cause a decreasing trend of backscattering during the growth phase. Typical representatives of these two types of crops are wheat for the first case and sunflower for the second one.
First of all, an accurate crop classification was performed in order to identify the various crop types responsible for the different backscatter behaviors. The backscattering trends have been simulated by using simple electromagnetic models based on radiative transfer theory. Subsequently, algorithms based on machine-learning approaches and in particular Neural Network methods have been implemented for estimating the crop biomass by using multi-frequency and multi-polarization SAR data at C and X band.
To this scope, an «experimental + model driven» approach was adopted. In detail, the ANN training was based on subsets of experimental data combined with model simulations, while testing and validation have been carried out using the remaining part of experimental data. This strategy preserved the statistical independence between training and validation sets, by also overcoming the site dependency of the data driven approaches based on experimental data only, thus ensuring some generalization capabilities of the proposed algorithms.
Although still preliminary, the results obtained are encouraging, confirming the peculiar sensitivity of each frequency to different vegetation features, and enabling the mapping of vegetation biomass in the test area with satisfactory accuracy.

Forests cover an estimated 31% of the Earth's global surface and therefore constitute a significant part of the biosphere. They fundamentally impact the carbon cycle as vegetation is able to absorb carbon atoms from the atmosphere and store them, by building up new biomass during their natural growth process.
Forests also majorly affect the local water-cycle, as the transpiration process redistributes ground-water into the atmosphere, impacting air temperature and weather in the process.
Forests are also critical for biodiversity preservation, an estimated 80% of all known terrestrial flora and fauna lives in them. Similarly, about 880 million people collect and produce fuel from wood while 90% of people living in extreme poverty have their livelihoods depending on forests.
To accurately estimate forest tree parameters such as canopy height (CHM) and above ground biomass (AGB), it is common practice to measure them manually on-site. This process can be both invasive, when individual trees are cut down to precisely assess their properties, or non-invasive, when a less intrusive approach is preferred over absolute accuracy.
The process is very expensive and time consuming, especially in remote areas. Therefore, in-situ measurement campaigns are feasible only for small surveys.
Airborne LiDAR systems also remain impractical and expensive when both large scale and low revisit time measurements are required, while spaceborne ones do not allow yet for the retrieval of wall-to-wall measurements.
As a consequence, spaceborne imaging systems for earth observation (EO) have gained wide interest in the last decades, as a large list of sensors and techniques is available delivering remote-sensing data at very large scales and low revisit times. Since this does not directly quantify forest parameters, it is necessary to model the relationship between the acquired data and the on-ground forest parameters.
Allometric equations are commonly used to indirectly relate forest parameters with RS data, but they require parameters to be tuned to the specific forest types and geographic locations to achieve good performance.
More sophisticated, physics-based modelling approaches have also been studied for the regression of forest parameters.
These tend to achieve high accuracy in their estimates, while retaining great spatial resolution.
To obtain these results, large amounts of data, auxiliary information or ground reference samples are required to invert the models.
With the recent advancements in machine learning and computer vision techniques, and the availability of large dataset collections from EO sensors, new approaches to forest parameter regression are starting to be explored.
Deep learning architectures have already found great success for classification tasks, as they analyze the spatial context information to generate higher level abstractions, producing features which typically possess a larger descriptive and discriminative content than both the input imagery and hand-crafted features.
On the other hand, comparatively little work still exists regarding the regression of physical and biophysical parameters from RS data, presumably due to the limited availability of large quantities of reference-data required for supervised training.
Aiming at providing large-scale, frequently updated CHM and AGB forest parameter metrics, our research effort focuses on overcoming the aforementioned limitations by proposing a multi-modal CNN-based regression framework, requiring only a single set of either single- or multi-source satellite imagery as input.
This multi-sensor approach represents a flexible solution for the continuous monitoring of forests when one or more input data sources are unavailable, and to otherwise achieve the best possible performance. In particular, we focus on combining high resolution Sentinel-2 optical imagery with TanDEM-X-derived interferometric SAR (InSAR) products, as they both provide fundamentally complementary information, and have been demonstrated to correlate well with forest parameters. The proposed data-driven multi-sensor approach consists in a deep multi-branch CNN architecture, where each of the modalities is associated to a separate feature extraction (encoder).
The spatial context extracted from these branches is then fused to supply a rich set of input features to a shared regression branch. We use a so-called cross-fusion approach to do this, which consists in a dedicated convolutional architecture that fuses different modalities through a set of convolutions and concatenations.
To assess the capabilities of the multi-branch architecture to fuse Sentinel-2 and TanDEM-X data and the regression performance of our framework, four tropical regions in Gabon, Africa have been considered. These correspond to reference data that has been acquired in the context of the 2016 AfriSAR campaign and consist of AGB maps, which have been derived at a ground sampling distance of 50m from airborne LiDAR measurements by fitting allometric equations on specifically acquired field-plot measurements.
We expanded the analysis period from mid 2015 to early 2017, since in 2016 only one Sentinel-2 satellite was available, which, combined with the extended cloud coverage over tropical regions, meant that only a small amount of imagery would have been available. We assumed that the changes in biomass are negligible within this time frame, as mainly tropical primary forest is considered.
During the learning phase, the network was trained on 32x32 pixel patches, using the mean square error (MSE) of the prediction as loss function for the backpropagation step. A validation set was used to select the best performing network across 10 training iterations. Finally, a separate test set was used to provide unbiased accuracy assessments.
Preliminary results in Gabon using Sentinel-2 optical and TanDEM-X interferometric SAR products are promising, showing agreement with the underlying assumptions and expectations. The root mean square error (RMSE) obtained on the test set is equal to 70.2 Mg/ha with a coefficient of determination R²=0.73, which is in line with the state-of-the-art methods.
We expect further optimization of the network and a more representative data set for training to further improve the estimation accuracy, setting the ground floor for the establishment of an effective tool for monitoring forest resources.

Fire danger is a description of the combination of both constant and variable factors that affect the initiation, spread, and ease of controlling a wildfire. The UK routinely experiences wildfires, typically with spring and mid/late summer peak occurrences, though winter wildfires do occur. In recent years, large-scale wildfire events in the UK have led to heightened concern about their behaviour and impacts (i.e. Saddleworth Moor and Winter Hill wildfires in 2018, Marsden Moor in February 2019). For instance, there were almost 260,000 wildfire incidents attended between 2009/10 and 2016/17 in England alone (avg. 32,000/year), requiring over 300,000 hours Fire and Rescue Services (FRS) attendance. In addition, the UK has an unusually complex fire regime which incorporates traditional management burning (Harper et al 2018), and episodic small to large-scale wildfires. While the largest wildfires (in terms of burned area) are on mountain, heath and bog (Forestry Commission England 2019), the largest number of wildfires occur in built-up areas, in particular in the rural-urban interface (RUI).

To assess, manage, and mitigate wildfire impacts, the likelihood of uncontrollable wildfires (Fire Danger) and the risk that they pose across the UK, must be quantified. Therefore, this project aims to establish and test the scientific underpinning and key components required to build an effective, tailored UK FDRS for use in establishing the likelihood and impact of current and future fire regimes. In order to accomplish this objective, we will: (i) produce UK fuel (i.e. flammable biomass) maps at the national, landscape and site-level, and to develop a site-level understanding of fuel structure; (ii) assess the moisture regimes in key fuel types across UK landscapes; (iii) determine flammability, energy content and ignitability of UK fuels to establish UK fuel models; (iv) determine the ranges of UK fire behaviour for key fuel types; (v) identify wildfire hotspots and with consideration of assets and communities at risk under current and future climate scenarios; and (vi) incorporate stakeholder knowledge and resources as an integral part of research delivery and impact generation.

In this presentation, we firstly provide an overview of the different components of the project, and secondly we explain in detail the techniques employed to map static fuel types over the UK for the year 2018. Fuels correspond to the vegetation classes with similar fire behaviour, represented as the biomass contributing to the spread, intensity and severity of the wildfires (Chuvieco et al 2003, Burgan et al., 1998). Our mapping fuel type is based on machine learning approaches that include different satellite data sources: Sentinel-1 and 2, Landsat-8 and ALOS PalSAR-2 to generate both height and above ground biomass (AGB) maps at national level at 10 m resolution. The national height map is based on all the available LiDAR data in the country provided by Digimap, and AGB national map is produced with the contribution of Forest Research and the National Forest Inventory. Resulting maps will be analysed with the UK Centre for Ecology and Hydrology Land Cover Map for 2018 to produce the UK national static fuel types map for 2018.

Understanding the hemiboreal forestland's role in the continental carbon cycle requires reliable quantification of their growing stock (forest biomass) at the regional scale. Remote sensing complements traditional field methods, enabling indirect fine-scale estimation of forest 3D structure parameters (primarily tree height) from high-density 3D point clouds by avoiding destructive sampling. In addition, carbon accounting programs and research efforts on climate-vegetation interactions have increased the demand for canopy height information, an essential parameter for predicting regional forest biomass [1]. Unfortunately, relatively high acquisition costs prevent airborne laser scanning (ALS), the most efficient and precise tool, from regularly mapping forest growing stock and dynamics. Therefore, in the last decade, there has been increasing interest in using very high resolution (ground sample distance (GSD) < 0.5 m) satellite-derived stereo imagery (VHRSI) to generate canopy height models (CHM) analogous to LiDAR point clouds to support forest inventory and monitoring. Despite the large offer of VHRSI sensors on the market (GeoEye, WorldView etc.), image-derived CHM performance for retrieving the forest inventory data in various geographical regions is still not fully understood [2]. However, while the ALS can penetrate the forest canopy and characterise the vertical distribution of vegetation, the VHRSI image-based point clouds only represent the non-transparent outer “canopy blanket” cover of dominant trees.

Thus, the present study assesses the potential of VHRSI sensors for an area-based prediction of growing stock (m3 ha-1) by deriving the main forest canopy height metrics from image-based point clouds and validating against the Latvian National Forest Inventory (NFI) data. The study area represents a typical hemiboreal forestland pattern across the eastern part of Latvia with predominantly mature, dense, closed-canopy evergreen pine, spruce and deciduous birch, black alder tree species.

The study workflow was divided into two stages. During the first stage, the study: (1) evaluated and compared the vertical accuracy and completeness of CHMs derived from airborne and VHRSI stereo imagery to reference LiDAR data; (2) analysed the differences in the CHM height estimates associated with different tree species; (3) examined the effect of sensor-to-target geometry (specifically base-to-height ratio) on matching performance and canopy height estimation accuracy [3]. As a result, the study confirmed the tendency for canopy height underestimation for all satellite-based models. The image-based CHMs of forests with dominated broadleaf species (e.g., birch and black alder) showed higher efficiency and accuracy in canopy height estimation and completeness than trees with a conical crown shape (e.g., pine and spruce). Furthermore, this research has shown that determining the optimum base-to-height (B/H) ratio is critical for canopy height estimation efficiency and completeness using image-based CHMs. This study found that stereo imagery with a B/H ratio of 0.2–0.3 (or convergence angle range 10°–15°) is optimal for image-based CHMs in closed-canopy hemiboreal forest areas.

At the second stage (currently being implemented), the study: (1) establish allometric relationships between field-derived (harvester data) individual tree volume and tree height; (2) use estimations from individual tree LiDAR measurements as training/reference data of growing stock for study area plots; (3) utilise a two-phase analysis that integrates both individual tree detection and area-based approaches (ABA) for precise forest growing stock estimation by using CHMs derived from airborne and VHRSI stereo imagery; (4) assesses the effect of ABA plot size on image-based CHM models performance and accuracy. The main goal of this study stage is to demonstrate that where field-plot (NFI) data are spatially limited, it is possible to use a hierarchical integration approach based on upscale forest growing stock estimates from individual trees to broader landscapes [4]. As for practical application and as an auxiliary tool for planning and managing forestry, the proposed method of mapping forest growing stock based on image-derived canopy height metrics will also be of great importance. However, compared to LiDAR, it is vital to remember that optical sensors are strongly influenced by solar illumination, sun-to-sensor and sensor-to-target geometry. The insufficient sunlight during the winter season, and summer season clouds, sometimes restrict the use of satellite sensors, making image-based vegetation monitoring problematic. The positive results of this study will facilitate Latvian regional forest growing stock inventories, monitoring and mapping by using VHRSI sensors as an adequate low-cost alternative to LiDAR data.



1. Fang, J.; Brown, S.; Tang, Y.; Nabuurs, G.-J.; Wang, X.; Shen, H. Overestimated Biomass Carbon Pools of the Northern mid- and High Latitude Forests. Clim. Change 2006, 74, 355–368, doi:10.1007/s10584-005-9028-8.
2. Fassnacht, F.E.; Mangold, D.; Schäfer, J.; Immitzer, M.; Kattenborn, T.; Koch, B.; Latifi, H. Estimating stand density, biomass and tree species from very high resolution stereo-imagery-towards an all-in-one sensor for forestry applications? Forestry 2017, 90, 613–631, doi:10.1093/forestry/cpx014.
3. Goldbergs, G. Impact of Base-to-Height Ratio on Canopy Height Estimation Accuracy of Hemiboreal Forest Tree Species by Using Satellite and Airborne Stereo Imagery. Remote Sens. 2021, 13, 2941, doi:10.3390/rs13152941.
4. Goldbergs, G.; Levick, S.R.; Lawes, M.; Edwards, A. Hierarchical integration of individual tree and area-based approaches for savanna biomass uncertainty estimation from airborne LiDAR. Remote Sens. Environ. 2018, 205, 141–150, doi:10.1016/j.rse.2017.11.010.

The primary science objective of ESA’s Climate Change Initiative Biomass project is the provision of global maps of above-ground biomass for four epochs (mid 1990s, 2010, 2017 and 2018) and, based on these, being capable of supporting the quantification of above-ground biomass change. Biomass in this context is given as above-ground forest biomass (AGB), which is defined following the FAO as the dry weight of live organic matter above the soil, including stem, stump, branches, bark, seeds and foliage woody matter per unit area, expressed in t/ha. AGB is also an Essential Climate Variable (ECV) within the Global Climate Observing System (GCOS).
Part of the project was holding a Biomass Change mapping workshop in late 2020. Due to the COVID-19 pandemic, the workshop was organized as virtual event 19th October – 6th November 2020. This virtual technical workshop enabled scientists from around the globe engaged in biomass change mapping to jointly formulate the underlying principles of forest biomass change estimation and the challenges connected with this and to develop meaningful estimates of the accuracy of such change measures.
Special circumstances - the pandemic - require special measures, in this case the virtual format of the workshop considering different time zones. The virtual workshop was running over a period of three weeks, where active participation was restricted to short periods within that timeframe. A domain was acquired and a dedicated website was created. A limited number of online presentations was made available at the first day of the workshop. They had to be watched prior to the live online-discussions concerning different topics (3 x 10 minutes per topic). Throughout the workshop further discussion were also possible via a dedicated discussion forum. To make participation possible for attendees from different parts of the world, all discussion rounds were hosted live in specific time slots. In this way, every interested participant of the workshop was given the opportunity to attend at least at one discussion round per week.
The first week was addressing issues related to “Defining and quantifying biomass change”. Three subtopics were selected: 1) The nature of change, 2) Change on the ground (e.g. linking traditional inventories with EO, standardisation of change descriptions and metrics, permanent plots with repeat coverage, biome-based allometric models), 3) Assessing the accuracy of AGB change estimates. The second week of the virtual workshop was handling questions related to “Biomass change from space” with the three subtopics 4) Change algorithms and methodologies, 5) Space and time considerations and 6) Validation of change.
During the workshop, a number of key questions concerning biomass change in general have been formulated jointly together:
How can global AGB be best mapped according to the controls on AGB change and to assess maximum biomass potential (consideration of climate, topography, latitude, flora and fauna)? And is this the best way to consider the different controls on biomass amounts (with respect to soils, air temperature, water resources, species distributions)?
Should biomass change be understood as relative to previously recorded amounts or maximum site potential amounts? This leads to the next question: Under which constraints is the detection of biomass change less relevant (e.g. in old growth forests or temporary biomass reduction by thinning) and how can we define threshold for saying that biomass change occurred?
How do we best describe the time-scales of biomass change ranging from rapid losses within a few days to weeks (deforestation, thinning) to yearly or decadal changes connected with forest growth?
This was leading to the overarching question: What is the best framework to use and follow for global biomass change classification?
All these issues are important for the continuation of the ESA CCI Biomass initiative. But the outcomes may also serve as a guideline for future fields of research and method development, which can be of use for the upcoming ESA P-band Biomass mission, now planned for launch in 2023.

Reed belts are an important subclass of aquatic vegetation as they represent some of the most important Blue Carbon ecosystems in the Baltic Sea. However, their extent has so far not been precisely mapped except in local field sampling experiments related to national inventory programmes. Differences in Normalized-Difference Vegetation Index (NDVI) have long been used as an indicator of vegetation in remote-sensed datasets and Earth Observation (EO). The differences are particularly large over coastal areas, where in the peak growth season (mid-to-late summer), uniformly vegetated areas such as reeds, sedges, rushes, and macrophytes have NDVI values around 0.7 ± 0.2 (1σ), whereas plain water has a relatively low NDVI, –0.2 ± 0.2 (1σ). In this work, Bayesian analysis is applied to identify areas of aquatic vegetation in monthly NDVI composites downloaded from the Sentinel-2 Global Mosaic (S2GM) service. These are then used as indicators for the occurrence of reed belts or other seasonal or permanent vegetation in coastal zones. The method is akin to naïve Bayes and outputs a value that is proportional to the probability of the pixel representing vegetation in water. The prior used is sensitive both to the NDVI and distance from shore; areas closer to the coastline are considered more likely to host aquatic vegetation. The method requires as its source datasets monthly composites of the NDVI from S2GM and a truthful sea mask, extractable from either national coastline layers or suitable land use classes from the Copernicus Coastal Zones data set.

The interpretation of aquatic vegetation has been carried out for the Finnish coast and two Swedish pilot areas in the south (Stockholm) and north (Piteå) in the context of project "Blue Carbon Habitats – a comprehensive mapping of Nordic salt marshes for estimating Blue Carbon storage potential –a pilot study", funded by the Nordic Council of Ministers. Training and test data were obtained from field-mapped reed outlines, and the probabilistic product was converted to a binary interpretation and sieved for too small areas or for areas too far from the shoreline. The ground truth of reed outlines aligns in general with the outlines inferred from EO, though the resolution (10 m) of the EO data limits the support near the shore. Unlike hypothesized, the posterior probability density of the Bayes product was not found to be strongly linked to species distribution nor the field-mapped reed belt density, and a different line of analysis will need to be carried out if these variables are to be predicted with remote-sensed observations.

Forest above-ground biomass (AGB) is identified as an essential climate variable (ECV) by the Global Climate Observing System (GCOS). Monitoring its spatial distribution and temporal variations is therefore a necessity to improve our understanding of climate change and increase our ability to predict its impacts.

In this study, we develop a novel approach to estimate AGB by using the TC×H variable, i.e. the product of percent tree cover (TC) and forest height (H) variables. To do so, we have used already available global datasets of TC and H. Percent tree cover is estimated from optical imagery, and we have retained the following products: a) the Global 2010 Tree Cover at 30m resolution derived from Landsat (Hansen et al., 2013) and b) the 2019 Tree Cover Fraction at 100m resolution derived from Proba-V (Buchhorn et al., 2020). Forest height is estimated from spaceborne lidar data and spatially extrapolated with optical imagery, and we have used the following products: a) the 2005 Global Forest Heights dataset at 1km resolution based on ICESAT-GLAS (Simard et al., 2011) and b) the 2019 Global Forest Canopy Height dataset at 30m resolution based on GEDI (Potapov et al., 2021). The spatial resolution of the datasets is degraded to 1km resolution to produce two TC×H layers: one for epoch 2005-2010 using the Hansen tree cover and the Simard height, and one for epoch 2019 using the Buchhorn tree cover and the Potapov height.

Relationships between TC×H and AGB were established using reference AGB estimates obtained from airborne Lidar datasets available within the ESA Climate Change Initiative Biomass project in the form of 100m resolution layers in Brazil, Indonesia, Australia, and the United States. The rationale behind the choice of the TC×H variable is that it constitutes a proxy of the vegetation volume, which itself is related to the AGB through the wood volumetric density. When the spatial resolution is degraded to 1 km, it is expected that the wood volumetric density can be considered almost uniform at the biome level. Therefore, we have aimed at establishing biome-specific AGB/TC×H relationships that are used to produce global estimates of AGB at 1km resolution for epochs 2005-2010 and 2019. These relationships are established through regressions based on a 3-parameter model, with the parameters estimated at each epoch (2005-2010 and 2019) and biome (temperate and boreal, wet tropical, and dry tropical). The inversion of these relationships provides global AGB estimates at 1km resolution at the two epochs. The AGB difference between the two epochs can be used to estimate the AGB change at a decadal scale.

This new approach can provide a low-cost and accurate alternative for the production of AGB maps at the kilometric scale. The validation of the AGB estimates is on-going and the first analysis results are promising. A quantitative comparison with the existing global AGB datasets (in particular the recently released CCI Biomass datasets) will be presented, in order to evaluate the strengths and weaknesses of each approach and identify the complementarity between methods.

Buchhorn, M., Smets, B., Bertels, L., De Roo, B., Lesiv, M., Tsendbazar, N.-E., Herold, M., Fritz, S., 2020. Copernicus Global Land Service: Land Cover 100m: collection 3: epoch 2019: Globe. https://doi.org/10.5281/zenodo.3939050

Hansen, M.C., Potapov, P.V., Moore, R., Hancher, M., Turubanova, S.A., Tyukavina, A., Thau, D., Stehman, S.V., Goetz, S.J., Loveland, T.R., Kommareddy, A., Egorov, A., Chini, L., Justice, C.O., Townshend, J.R.G., 2013. High-Resolution Global Maps of 21st-Century Forest Cover Change. Science 342, 850–853. https://doi.org/10.1126/science.1244693

Potapov, P., Li, X., Hernandez-Serna, A., Tyukavina, A., Hansen, M.C., Kommareddy, A., Pickens, A., Turubanova, S., Tang, H., Silva, C.E., Armston, J., Dubayah, R., Blair, J.B., Hofton, M., 2021. Mapping global forest canopy height through integration of GEDI and Landsat data. Remote Sens. Environ. 253, 112165. https://doi.org/10.1016/j.rse.2020.112165

Simard, M., Pinto, N., Fisher, J.B., Baccini, A., 2011. Mapping forest canopy height globally with spaceborne lidar. J. Geophys. Res. Biogeosciences 116. https://doi.org/10.1029/2011JG001708

The key parameters provided by the Soil Moisture and Ocean Salinity (SMOS) mission over land are soil moisture (SM) and L-band vegetation optical depth (L-VOD). Although the retrieval of SM was the low hanging fruit of the mission, the information about vegetation has reached maturity causing a growing interest in testing the L-VOD product and using it in applications. Previous studies investigated the correlation between L-VOD and vegetation properties, such as vegetation height and forest biomass, made available by data bases.
In this paper, L-band vegetation optical depth (L-VOD) retrieved by SMOS is compared against vegetation parameters (RH100 and PAI) retrieved by Global Ecosystem Dynamics Investigation (GEDI) lidar instrument, recently launched by NASA. L-VOD was retrieved using the recent v700 version of SMOS level 2 algorithm. In order to manage the different spatial resolutions, GEDI parameters were averaged within SMOS pixels and a threshold to the minimum number of GEDI samples per SMOS pixel was applied. The investigation is multitemporal, since spatial correlations between monthly averages are investigated from May 2019 to April 2020, and a temporal extension to a two year interval is in progress.
The analysis was initially done for four large continents. For Africa and South America, mostly covered by tropical vegetation, the Pearson correlation coefficients between L-VOD and RH100 are higher than 0.8 in all months of the year. Conversely, seasonal effects are observed in North America and Asia, producing a lower correlation coefficient in colder months. RMS differences between L-VOD’s retrieved by SMOS and the ones obtained using a linear regression over RH100 are lower than 0.2 for all cases, and close to 0.1 for most cases. Using PAI in place of RH100 slightly lower spatial correlations are generally achieved.
The analysis was repeated considering three latitude belts: Northern, Tropical, and Southern. In the tropical belt the coefficients of L-VOD versus RH100 regression are stable and the Pearson correlation coefficient is higher than 0.88 for all months of the year. For Northern vegetation the regression slope and the Pearson correlation coefficient are stable from May to September, but decrease in the winter season. Lower Pearson correlation coefficients (about 0.7) are found in the Southern belt, due to reduced dynamic ranges of L-VOD and vegetation height.
All correlation coefficients between v700 L-VOD and RH100 are better with respect to L-VOD from previous level 2 versions. Overall, the obtained results confirm the good potential of L-VOD to monitor vegetation height in different environments. The synergic use of GEDI and SMOS L-VOD data sets can improve the accuracy and/or the timeliness in monitoring vegetation changes occurring at yearly or monthly time scales, such as deforestation, re-growth and desertification.

Passive microwave observations from 1.4 to 36 GHz already showed sensitivity to vegetation parameters, primarily through the calculations of the Vegetation Optical Depth (VOD) at individual window frequencies, separately. Here we evaluate the synergy of this frequency range for vegetation characterization over Tropical forest, through the estimation of two vegetation parameters, its foliage and the photosynthesis activity as described by the Normalized Difference Vegetation Index (NDVI), and its woody components and carbon stock as described by the Above Ground Carbon (AGC), using different combinations of channels in the considered frequency range. Neural network retrievals are trained on these two vegetation parameters (NDVI and AGC), for several microwave channel combinations, including the future Copernicus Imaging Microwave Radiometer (CIMR) that will observe simultaneously in window channels from 1.4 to 36 GHz, for the first time. This methodology avoids the use of any assumptions in the complex interaction between the surface (vegetation and soil) and the radiation, as well as any ancillary observations, to propose a genuine and objective evaluation of the information content of the passive microwave frequencies for vegetation characterization. Our analysis quantifies the synergy of the microwave frequencies from 1.4 to 36 GHz. For the retrieval of NDVI, the coefficient of determination R2 between retrieved and true NDVI reaches 0.84 when using the full 1.4 to 36 GHz range as will be measured by CIMR, with a retrieval error of 0.07. For the retrieval of AGC, the coefficient of determination R2 reaches 0.82 with CIMR, with an error of 21 Mg/ha. This study also confirmed that 1.4 GHz observations have the highest sensitivity to AGC, as compared to other frequencies up to 36 GHz, at least under tropical environments.
CIMR will provide valuable ecological indicators to enhance our present global vegetation understanding. Considering both vegetation aspects together (foliage photosynthesis activity and carbon stocks) offers a more robust and consistent characterization and assessment of long-term vegetation dynamics at large scale. CIMR will operate in synergy with MetOp-SG that carries the ASCAT scatterometer at 5.2 GHz. The complementarity between CIMR and the active microwave observations from ASCAT will also be evaluated, over Tropical forests, for vegetation characterization.

Blockchain Applications for Biomass Measurement and Deforestation Mitigation

Accurate estimation of forest above-ground biomass and its change over time is critical to forest conservation efforts and the consequence of the current voluntary carbon market. A positive change in biomass through planting more trees is important, but it must be noted that afforestation is merely the first step in producing an increase in global carbon sequestration. The true variable influencing the outcome of these efforts is the resilience of the biomass, and tree growth that resilience results in. Typically, forest densities range from 1000 to 2500 trees per hectare and have a carbon sequestration rate of up to about 10 tonnes per hectare per year in the tropics (0.8 to 2.4 tonnes in boreal forests, 0.7 to 7.5 tonnes in temperate regions and 3.2 to 10 tonnes in the tropics). By the age of 100, one broadleaf tree (commonly found in the tropical rainforests) could have sequestered up to one tonne of carbon. In comparison, chopping and burning just about 5 to 10 average-sized pine trees (~450 Kg dry weight each) instantly releases back all of the carbon that one hectare of trees captured in one year. This points to an extreme negation effect that logging can have, and can be a powerful point of change for policy makers.

The new forests that have been planted in the past two decades represent merely 5% of the net global carbon sink. While these numbers will grow with time, it is currently far more important to monitor and prevent deforestation of existing mature forests. It is also potentially one of the most difficult interventions to implement. Most deforestation and forest degradation are concentrated in the tropics because of both illegal logging activities that can go undetected due to the small spatial scales they occur at, and because of general forest clearing for cattle pasture and agricultural expansion. The most drastic effects of this rampant deforestation have been witnessed in the Amazon rainforest, which, at the time of writing this, has already / is strongly tending towards becoming a net emitter of carbon instead of a net sink, if the given rates of deforestation continue. Since the turn of the century, Brazil alone has released 32.5Gt of CO2 from deforestation and for reference, the average amount of CO2 annually emitted across the globe varies around 40 Gt. The rate of carbon emission also varies according to forest type. Old-growth primary forests, unlike their secondary and fast-rotation counterparts, can release carbon that has taken centuries to get stored. Hence, illegal logging, especially deep within primary forests needs early detection through regularly updated AGB change estimation.

At a policy level, accurate and timely detection of changes in biomass can be of value to countries that are attempting to recognize indigenous peoples and local communities as owners of their lands. Enforcing the rights of indigenous communities is a proven strategy to protect standing forests and enhance the carbon stored in them.

Governments across the globe have been actively incorporating forest landscape restoration measures in their policies. However, the effectiveness of these interventions towards carbon removal and climate change mitigation is difficult to quantify, especially in regions where there is insufficient biomass data. To fill these gaps in knowledge, various programs have provided open-source access to sensor-fused datasets at resolutions varying between 100m (ESA Climate Change Initiative) to 500m (NASA Pantropical AGB dataset).

Our work aims to utilize machine learning techniques such as Random Forest and XGBoost to train our algorithm on recently developed AGB datasets and Vegetation Indices extracted from satellite image radiances. The correlation between these inputs is then used to predict AGB in a different location and year. Since above-ground biomass accounts for about 27% of the entire carbon sequestered by a tree, the mathematical difference between the pixel values of two independent AGB predictions over simultaneous years allows us to estimate the total carbon sequestration at that pixel, in that year. Moreover, biomass change detection can help clearly identify logging activity, storm damage, restoration after forest fires and reforestation efforts being undertaken by forest managers. This allows for monitoring of policy implementations as well. Through sensor fusion with multiple satellite datastreams including SAR data, we can monitor large scale regions (including remote and inaccessible ones) at night, and through clouds (a major issue in imaging the tropics), with a rapid revisit rate.

Predictions of biomass change and carbon sequestration both occur at the pixel resolution of the training dataset, although we are also working on methods to increase pixel resolutions of the final products through the use of deep neural networks. The ability to monitor biomass change at 100m resolution and lower, will also assist with forest boundary change measurement. Forest boundaries can be substantial areas of change in forest expanse due to easy access, and estimating those changes can be very helpful to forest managers.. Moreover, we test several vegetation index combinations to better understand and potentially provide insight towards standardizing best practices for AGB prediction models globally.

Our work further contributes to solving the issue of the severe lack of multi-temporal AGB datasets. Projects such as the NASA GEDI mission, while providing very high-resolution AGB estimates, are currently only a one-time estimate. A similar problem occurs with the ESA CCI biomass datasets which are only valid for 2010, 2017, and 2018. Internally consistent AGB change datasets were only made available in December of 2021. Other datasets are also only available for one year randomly placed throughout the past two decades. This points to the fact that while work is being consistently done to acquire and formulate these datasets, there is a need for predictive software to estimate continuous time series of AGB change across several years, which can then be validated by ground truth and reference datasets whenever they become available.

Environmental protection does, however, come at the cost of economic growth; and this is a major hurdle, especially in developing nations. Therefore, a highly effective way of incentivizing countries to strictly control deforestation is to provide them monetary compensation through the use of carbon credits.

The carbon credit market has a key role to play in the solution to the problem of climate change and dClimate is entering this field using machine learning mechanisms and advanced AI algorithms. The current voluntary carbon market does not put enough emphasis on preventing deforestation. Only 32% of carbon offsets deal with preventing deforestation, and the IPCC believes the number of carbon removal projects must increase in order to limit warming to 1.5°C.

One of the primary issues with carbon offsets is the lack of transparency in the market. dClimate is revolutionizing the industry by verifying our own offsets using blockchain technology, which will allow buyers and project creators to view a transparent immutable ledger with all needed information available. To do this, we are creating an above-ground biomass estimation and monitoring system, which will allow us to price tokens based on the amount of carbon sequestered by existing biomass, and stored in the form of AGB change. Currently, the VCM registries are plagued with double or triple counting wherein the same parcel of land is sold multiple times. This not only prevents adequate market dynamics for the price of carbon offsets, but also limits the growth of the industry. In addition, the measurement of deforestation and carbon output is traditionally non-standard as it entails very bespoke methodologies which are not equipped to handle the problem at scale.

Leveraging the intersection of the simultaneous advances in many decentralized technologies such as Chainlink, IPFS, and distributed ledgers (Ethereum) we are able to create new financial ReFi (regenerative finance) primitives which facilitate carbon price discovery. Additionally, through the use of decentralized execution environments, all computations are transparent and can be inspected by anyone providing a platform for trustless interoperability without having to rely on centralized failure points. By creating this infrastructure we not only create the tools to mitigate deforestation, but also accurately measure other parts of the collective climate economy.

As a result of bringing together the new multi-resolution (spatial and temporal) datasets from multiple global organizations, increasing end-to-end transparency, and creating pressure on countries and stakeholders through a penalty system for anthropogenic biomass reduction, our work will develop detailed maps of Above Ground Biomass and its spatio-temporal variation. This will not only provide financial impetus to all nations who choose to use our services (especially to low-income countries with high biomass reserves), but will also assist the global scientific community by providing a rapidly updated database through an easily accessible API, aimed at creating a standard system of carbon emission control and sequestration measurement.

Since the collapse of the Soviet Union and being in transition to a new forest inventory system, Russia has reported almost no change in growing stock (+1.3%) and biomass (+0.6%). The Forest and Agriculture Organization of the United Nations (FAO) Forest Resources Assessment (FRA) national report 2020 presented 81.1 billion m3 of the growing stock volume (GSV) or 63.0 billion tons in above ground biomass (73.3 t/ha). FAO FRA national report is based on outdated State Forest Register. The first cycle of National Forest Inventory (NFI) was accomplished in Russia in 2020. The results of the new NFI were announced at the UN Climate Change Conference of the Parties (COP26) in Glasgow. The total GSV of Russian forest is 111.7 billion m3, or 38% higher than in the FAO FRA report. This discrepancy explained by the transition to a new inventory system – NFI and the gap in updating forest information.

In Russia, the long intervals between consecutive surveys and the difficulty in accessing very remote regions in a timely manner by an inventory system make satellite remote sensing (RS) an essential tool for capturing forest dynamics and providing a comprehensive, wall-to-wall perspective on biomass distribution. However, observations from current RS sensors are not suited for producing accurate biomass estimates unless the estimation method is calibrated with a dense network of measurements from ground surveys (Chave et al., 2019). Here we calibrated models relating two global RS biomass data products (GlobBiomass GSV (Santoro, 2018) and CCI Biomass GSV (Santoro & Cartus, 2019)) and additional RS data layers (forest cover mask (Schepaschenko et al., 2015), the Copernicus Global Land Cover CGLS‐LC100 product (Buchhorn et al., 2019)) with ca 10,000 ground plots to reduce nuances in the individual input maps due to imperfections in the RS data and approximations in the retrieval procedure (Santoro, 2019; Santoro et al., 2021). The combination of these two sources of information, i.e., ground measurements and RS, utilizes the advantages of both sources in terms of: (i) highly accurate ground measurements and (ii) the spatially comprehensive coverage of RS products and methods. The amount of ground plots currently available may be insufficient for providing an accurate estimate of GSV for the country when used alone, but they are the key to obtaining unbiased estimates when used to calibrate RS datasets (Næsset et al., 2020).

Our estimate of the Russian forest GSV is 111±1.3 billion m3 for the official forested area (713.1 million ha) for the year 2014, which is very close to the NFI aggregated results. An additional 7.1 billion m3 were found due to the larger forested area (+45.7 million ha) recognized by RS (Schepaschenko et al., 2015), following the expansion of forests to the north (Schaphoff et al., 2016), to higher elevations, in abandoned arable land (Lesiv et al., 2018), as well as the inclusion of parks, gardens and other trees outside of forest, which were not counted as forest in the State Forest Register. Based on cross-validation, our estimate at the province level is unbiased. The standard error varied from 0.6 to 17.6% depending on the province. The median error was 1.6%, while the area weighted error was 1.2%. The predicted GSV with associated uncertainties is available here (https://doi.org/10.5281/zenodo.3981198) as a GeoTiff at a spatial resolution of 3.2 arc sec. (ca 0.5 ha).

Acknowledgements
This study was partly supported by the European Space Agency via projects IFBN (4000114425/15/NL/FF/gp). The NFI data preparation and pre-processing were financially supported by the Russian Science Foundation (project no. 19-77-30015). FOS data preparation and processing for the Central Siberia were supported by the RSF (project no 21-46-07002).

References
Buchhorn, M., Bertels, L., Smets, B., Lesiv, M., & Tsendbazar, N.-E. (2019). Copernicus Global Land Service: Land Cover 100m: version 2 Globe 2015: Algorithm Theoretical Basis Document. Zenodo. https://doi.org/10.5281/zenodo.3606446
Chave, J., Davies, S. J., Phillips, O. L., et al. (2019). Ground Data are Essential for Biomass Remote Sensing Missions. Surveys in Geophysics, 40(4), 863–880. https://doi.org/10.1007/s10712-019-09528-w
Lesiv, M., Schepaschenko, D., Moltchanova, E., et al. (2018). Spatial distribution of arable and abandoned land across former Soviet Union countries. Scientific Data, 5, 180056. https://doi.org/10.1038/sdata.2018.56
Næsset, E., McRoberts, R. E., Pekkarinen, A., et al. (2020). Use of local and global maps of forest canopy height and aboveground biomass to enhance local estimates of biomass in miombo woodlands in Tanzania. International Journal of Applied Earth Observation and Geoinformation, 102138. https://doi.org/10.1016/j.jag.2020.102138
Santoro, M. (2018). GlobBiomass—Global datasets of forest biomass [Data set]. https://doi.org/10.1594/PANGAEA.894711
Santoro, M. (2019). CCI Biomass Product User Guide (p. 35). GAMMA Remote Sensing. https://climate.esa.int/sites/default/files/biomass_D4.3_Product_User_Guide_V1.0.pdf
Santoro, M., & Cartus, O. (2019). ESA Biomass Climate Change Initiative (Biomass_cci): Global datasets of forest above-ground biomass for the year 2017, v1 [Application/xml]. Centre for Environmental Data Analysis (CEDA). https://doi.org/10.5285/BEDC59F37C9545C981A839EB552E4084
Santoro, M., Cartus, O., Carvalhais, N., et al. (2021). The global forest above-ground biomass pool for 2010 estimated from high-resolution satellite observations. Earth System Science Data, 13, 3927–3950. https://doi.org/10.5194/essd-13-3927-2021
Schaphoff, S., Reyer, C. P. O., Schepaschenko, D., Gerten, D., & Shvidenko, A. (2016). Tamm Review: Observed and projected climate change impacts on Russia’s forests and its carbon balance. Forest Ecology and Management, 361, 432–444. https://doi.org/10.1016/j.foreco.2015.11.043
Schepaschenko, D., Shvidenko, A. Z., Lesiv, M. Yu., et al. (2015). Estimation of forest area and its dynamics in Russia based on synthesis of remote sensing products. Contemporary Problems of Ecology, 8(7), 811–817. https://doi.org/10.1134/S1995425515070136

Forage provision is an important indicator of rangeland health and is reliable for evaluating land degradation. In dry rangelands it is largely limited by moisture availability compounded with grazing pressure as it sustains a significant proportion of livestock-based systems. For sustainable and adaptive management, parameters such as biomass production and forage quality are of key interest. Yet, their quantification and monitoring still remains laborious and costly. Advancing remote sensing technologies such as hyperspectral readings and drone imaging enable rapid, repeatable and non-destructive estimations of these parameters that can be applied over large spatial scales. While these are increasingly being integrated for ecological research, robust prediction models supported by field data is still lacking, especially in highly dynamic systems like semi-arid savannahs. In our study we aim to answer the following research questions: (1) to what extend can we model forage provision (quality and quantity) from resampled hyperspectral data? (2) Can we model forage provision from UAV-based multispectral imagery calibrated with field spectrometer prediction models? (3) How do artificial hyperspectral data, interpolated from multispectral data enhance the prediction quality? (4) How does forage provision vary between two differently managed rangelands? To address these questions, we took hyperspectral readings with a field spectrometer from herbaceous canopies along transects in two management types in a Namibian semi-arid savannah. Plant biomass samples were collected at the reading areas to measure forage quantity and forage quality. Machine learning and deep learning methods were used to establish hyperspectral prediction models for both forage quality and quantity. We applied these models to hyperspectral readings from a broader area. For upscaling the hyperspectral models, we acquired drone multispectral imagery along the same transects. Multispectral prediction models were set up using the predicted values from the hyperspectral prediction model. As predictors for the model we used the pure spectra, derived vegetation indices and artificial hyperspectral data from interpolating the multispectral bands. We then created forage quantity and quality maps to visualize and compare forage provision dynamics in the two management systems. While field-based hyperspectral models offer greater spectral resolution for assessing complex forage quality parameters, and drone imagery offer unprecedented spatial and temporal data products for mapping forage parameters at a landscape level, independently they are limited. Thus, emerging UAV-based hyperspectral imagery minimizes these discrepancies, a technology that will catapult remote sensing to map even more complex variables and resolve ecological questions.

Agriculture is a critical source of employment in rural Colombia and is one of the sectors most affected by climate and climate change and where solutions to key challenges affecting the productivity and sustainability of forages and the livestock sector are required. Increasing yields of forage crops can help improve availability and affordability of livestock products while also easing pressure on land resources through enhanced resource utilisation. This study aims to develop remote sensing-based approaches for forage monitoring and biomass prediction at local and regional levels in Colombia. Local access to such information can help improve decision making and increase productivity and competitiveness while minimising impacts on the environment. Ten locations were sampled between 2018 and 2021 across climatically distinct areas in Colombia, comprising five farms in Patía in Cauca department, four farms in Antioquia department, and one research farm at Palmira in Valle de Cauca department. Ash content (Ash), crude protein (CP %), dry matter content (DM g per square meter) and in-vitro digestibility (IVD %) were measured from different Kikuyu and Brachiaria grasses during the field sampling campaigns. Multispectral bands from coincident Planetscope acquisitions along with various derived vegetation indices (VIs) were used as predictors in the model development. To determine the optimum models, the improvement capabilities of using an averaging kernel, feature selection approaches, various regression algorithms and metalearners (simple ensembling and stacks) were explored. Several of the applied algorithms have built-in best feature selection functions so to test model improvement capabilities of an independent feature selection approach for algorithms that have one built-in and for those that do not, all models were run a) with no feature pre-selection, b) with Recursive Feature Elimination (RFE, package: caret) and c) with Boruta (package: Boruta) feature selection. A range of algorithms (n=26) belonging to classes of decision trees, Support Vector Machines, Neural Networks, distance-based methods, and linear approaches were tested. All algorithms including metalearners were tested with each of the three feature selection approaches while employing 10-fold cross-validation with 3 repeats. In the performance evaluation based on unseen test data, CP and DM was predicted relatively well for all three sites (R2 0.52 – 0.75, RMSE 1.7 – 2.2 % and R2 0.47 – 0.65, RMSE 260 – 112 g/m2 respectively). As part of the study, the investigation was carried out in cooperation with smallholder farmers to determine their attitudes and potential constraints to mainstreaming such technologies and their outcomes on the ground. Through improving communication between earth observation and agricultural communities and the successful integration of satellite-based technologies, future strategies can be implemented for increasing production and improving forage management while maintaining ecosystem attributes and services across tropical grasslands.

The rise in atmospheric CO2 due to anthropogenic emissions is the leading cause of climate change. In order to avoid reaching tipping points in the Earth System, efforts to cut down emissions and compensate for existing atmospheric CO2 are pursued. Within this context, nature-based solutions such as reforestation, afforestation and agroforestry favour the potential of carbon sequestration through tree growth and health. Though, the kick-off of planting activities is insufficient for these initiatives to present a substantial impact. Routine field maintenance and restoring the ecosystems are capital. Since several decades are necessary before observing the positive impacts of such initiatives, long-run investments become indispensable, and with them, monitoring and metrics. Current methods that measure carbon storage in forests and their evolution are based regionally on Forest National Inventories and globally on products derived from satellite imagery at coarse resolution. These solutions frequently lack the temporal and spatial coverage to ensure traceability and transparency to monitor most interventions.

In order to follow the evolution of local and regional scale nature-based projects effectively, a monitoring strategy relying on remote sensing is presented, covering the needs of new afforested sites and reforestation and agroforestry management in existing forested areas. The methodology revolves around skilled monitoring of Above-Ground Biomass (AGB) -one of the most reliable means to assess natural carbon sinks. The proposed monitoring system covers regional stakeholders' needs to trigger payments for the environmental services implemented by over a thousand local farmers.

At high resolution, several reforestation efforts in the Sahel area in Africa are being monitored using Very High Resolution (VHR) imagery from Airbus' Pleiades mission. The detection of individual trees is possible thanks to the mission's pan-sharpened resolution of 0.5m. A monitoring system for larger-scale regional projects using medium-resolution imagery of 20m pixels is also presented. For the latter, data sources include Copernicus Sentinel-1 and Sentinel-2 missions for Synthetic Aperture Radar (SAR) and Multi-Spectral imagery, respectively, LiDAR-based AGB data provided by the Global Ecosystem Dynamics Investigation (GEDI) mission on board of the International Space Station (ISS), Land Cover maps and Digital Elevation Models (DEM).

In order to train, test and validate regression methods that predict the evolution of carbon stored in individual trees, reliable and standardised measurements at tree level are necessary. A dedicated in-situ survey strategy has been designed collaboratively with local communities and field experts in the Sahel to overcome the limitations of horizontal GNSS resolution and obtain reliable measurements at the tree level.

While the in-situ surveying is not taking place due to security constraints in West Africa, a preliminary study is carried out in Catalunya, Spain, benefiting from an AGB dataset obtained from the Spanish 4th National Forest Inventory (NFI-4). This proof of concept shows the correlations of the individual data sources with field biomass and the combined use of all the datasets in the methodology to address biomass assessment. The study over the region of Catalunya serves as a basis to transfer the methodology to the Sahel region, where the aforementioned nature-based projects are taking place.

Further to the CO2 sequestration potential, the beneficial side-effects of nature-based solutions include improved soil quality, increased crop yield, ground temperature and biodiversity recovery, and positive socio-economic impacts, which are rarely quantified. Additional metrics are presented as valuable information to the overall Key Performance Indicators to add a comprehensive vision of the reforestation and afforestation activities on the local communities involved.

The presented approach is developed within the JESAC project (https://www.jesac-project.com), integrating a virtual monitoring platform. Payments for Environmental Services (PES) will be triggered once the trees have stored a certain amount of carbon. These payments cover in-field activities for land restoration and voluntary carbon offsetting, which is traced transparently through blockchain technology.

Several studies have highlighted the saturation effects of L-Band SAR signal sensitivity at the increasing of forest density. In those cases, a direct modeling approach or an empirical regression guided by ground sampled measurements can be not effective to estimate the Above Ground Biomass (AGB) values higher than ~ 150 – 200 t/ha. Machine learning approaches are therefore proposed in recent literature to deal with these types of constrain with active (and passive) microwave monitoring of forest, by including different type of ancillary information.
In the ESA MAFIS project we have teste the feasibility of a Random Forest (RF) procedure including SAR and optical data. The strength of the RF solution consists in the possibility to include different type of Earth observation quantities in addition to the L band backscatter for characterizing the AGB. In this way the L band SAR signal is coupled with multispectral optical indexes, for limiting the saturation effects of the SAR signal, without explicitly dealing with the complex non-linearity of the combination of the input variables. Anyway this hypothesis can be effectively exploited only if a sufficient set of reference data for the AGB are available. In general, the use of in situ measurements sampled on tens to hundreds grounds plots cannot be sufficient to proper finalize the training phase of a data drive algorithm. In the MAFIS project we tried to overcame such limitation by exploiting recent aerial LiDAR data made available for the Veneto Region over the alpine areas of Lorenzago di Cadore and Bosco del Cansiglio (Noth-East of Italy). Those areas were affected by the Vaia storm, occurred from the 26th to the 30th October 2018. This event caused a dramatic loss of forest area in different Italian regions due to strong winds that pulled down a massive quantitative of trees. Regione Veneto acquired a large set of aerial LiDAR data after the Vaia storm for mapping the extension of the affected areas. This quite unique dataset represent a good opportunity to evaluate the effectiveness of a Random Forest approach for the AGB retrieval by means of the fusion of L-band SAR data and multispectral data. In fact the forest areas acquired during the flights spans for several tens of hectares and provides thousands of training examples of intact areas over the forest AGB can be derived from the LiDAR measurements of tree’s height. In particular, the LiDAR data have been processed to derive the Digital Terrain model (DTM) and the Digital Surface model (DSM). These latter have been used to derive the tree height layer over the forest considered areas. Finally, a corrected version (fitted to several local data acquired during the MAFIS project in situ survey) of dendrometric tables of the second Italian National Forest Inventory (INFC), which define volume estimation equations adapted to the different forest species, have been applied to the most common tree species of the considered Alp regions, i.e. Fagus, Abies Alba and Larix Decidua, for computing the LiDAR based AGB layer, which ranges between ~200 and ~1000 m3/ha over the analysed regions. The latter has been the divided in the training and test set, used respectively to train the RF model and to test its performances.
The input data to the RF model are the HH and HV backscattering coefficients, extracted from ascending and descending SAR ALOS-2 PALSAR-2 L1.1 products, and multispectral reflectance (in VIS, NIR an SWIR), extracted from Sentinel-2 L2A products. Both the ALOS-2 and the Sentinel-2 data have been collected on specific dates comparable with the time range of the aerial LiDAR acquisitions.
The results of the trained RF model evaluated over the independent test set shown very encouraging results, with a correlation coefficient higher than 70% and reporting very coherent behavior of the spatial patterns of AGB within the mountain landscape. Finally, the insurgence of the saturation effects is registered for a threshold of about 900 m3/ha.

Reducing uncertainty in the estimation of aboveground biomass (AGB) stocks is required to map global aboveground carbon stocks at high spatial resolution (< 1 km) and monitor patterns of woody vegetation growth and mortality to assess the impacts of natural and anthropogenic perturbations to ecosystem dynamics. The NASA Global Ecosystem Dynamics Investigation (GEDI) is a lidar mission launched by NASA to the International Space Station in 2018 that has now been collecting science data since April 2019 and is expected to continue to at least January 2023. These observations underpin efforts by the NASA Carbon Monitoring System (CMS) to advance pantropical mapping of forest and woodland AGB and AGB change through fusion of GEDI with interferometric Synthetic Aperture Radar (InSAR) observations from current and upcoming missions. These aim to facilitate much needed improvements to national-scale carbon accounting and other monitoring, reporting and verification (MRV) activities across forest and woodland ecosystems in pantropical countries. Here we present a novel fusion approach that combines billions of GEDI measurements with high resolution InSAR data acquired between 2010 and 2019 by TanDEM-X, resulting in wall-to-wall canopy height and AGB estimates at 1 ha spatial resolution across the pantropics, including Brazil, Gabon, Mexico, Australia. We first present AGB prediction models that use GEDI measurements of canopy height and cover at the scale of field plots typically used for calibration and validation of satellite mapping of AGB. These include the footprint scale (0.0625 ha) and, through aggregation at International Space Station (ISS) orbital crossovers, the 1 and 4 ha scales specified by upcoming spaceborne InSAR missions designed for global mapping of AGB (NASA/ISRO NISAR, ESA BIOMASS). We show that the addition of GEDI measurements improved 1 ha TanDEM-X canopy height RMSE by 16.6-38.2% over pilot countries and reduced the magnitude of systematic deviations observed using TanDEM-X alone. Finally, using new models that link GEDI plot scale estimates of AGB with vertical and horizontal canopy structure metrics from TanDEM-X, and Generalized Hierarchical Model-Based inference (GHMB) to propagate uncertainty, we compare the precision of estimates achieved through our fusion approach to those achieved using GEDI or TanDEM-X alone. This study defines good practices for linking GEDI observations with those from satellite imaging SAR that are based on refined measures of quality and geolocation, and their impact on estimates of AGB uncertainty achieved through fusion of GEDI with satellite InSAR. Our approach takes full advantage of more direct estimates of structure and AGB from GEDI, and further highlights the importance of a formal and transparent framework to estimate uncertainty and enable the separation of true and spurious change in the monitoring of AGB across pantropical forest and woodland ecosystems.

Earth observation is a necessary resource in understanding some of the world’s most sensitive ecosystems. Kenya’s coastal communities have suffered greatly from land degradation and poor soil health due to climate change and over farming. This project aims to look deeper into the ways that we can save these rural communities by using satellite imagery to get a better understanding of the Green World Campaign’s regenerative efforts throughout coastal Kenya. Using very high resolution (VHR) imagery from MAXAR’s Worldview satellites, the intent is to understand how this extremely high resolution imagery coupled with field data and random forest classification methods can help to identify a more accurate understanding of soil heath and tree growth, thus directly impacting the future livelihood of these communities.
In conjunction with the ever-evolving high resolution imagery and SmallSat constellation expansion, we propose a conceptual model for both the public and private sectors that marries accurate data with direct funding opportunities using cryptocurrencies through biomass and carbon monitoring practices. This uniquely holistic model has proven it is capable of restoring the economy and ecology of communities struggling on the front lines of climate change. This regenerative model's “people-and-planet” approach addresses the health of both landscapes and communities, leading to improved rural livelihoods, nutrition, biodiversity, soil health, and carbon “drawdown". Earth Observation plays a critical role in this process, for both understanding the past landscape's soil levels as well as looking toward future imagery analysis. Having the ability to visualize this landscape change in real-time is a direct confirmation of progress on both the micro and macro levels of climate resilience. Increasing the effectiveness of studying remote regions will not only be of importance to rural communities in Kenya but will also be usable in other remote areas of the world, helping to gain a greater perspective of global system change and forest abundance.

Vegetation biomass is a globally important climate-relevant terrestrial carbon pool. In tundra permafrost lowland landscapes North of the treeline, the vegetation low-level structure poses a challenge for the derivation of the plant biomass both, from optical and SAR satellite remote sensing. Still, a range of tundra types have some spectral or structural characteristics for land cover classification. Higher vegetation, such as high-growing shrubs occur in small patch sizes. In this study we investigate to which extent data from Sentinel-2 and Sentinel-1 missions provide a landscape-level opportunity to upscale tundra vegetation communities and biomass for high latitude terrestrial environments.
We assessed the applicability of landscape-level remote sensing for the low Arctic Lena Delta region in Northern Yakutia, Siberia, Russia. The Lena Delta is the largest delta in the Arctic and is located North of the treeline and the 10 °C July isotherm at 72° Northern Latitude in the Laptev Sea region. Vegetation and biomass field data from Elementary Sampling Units ESUs (30 m x 30 m plot size) and shrub samples for dendrology were collected during a Russian-German expedition in summer 2018 in the central Lena Delta.
We evaluated circum-Arctic harmonized ESA GlobPermafrost land cover and vegetation height remote sensing products covering subarctic to Arctic land cover types for the central Lena Delta. The products are freely available and published in the PANGAEA data repository under https://doi.org/10.1594/PANGAEA.897916, and https://doi.org/10.1594/PANGAEA.897045.
We also produced a regionally optimized land cover classification for the central Lena Delta based on the in-situ vegetation data and a summer 2018 Sentinel-2 acquisition that we optimized on the biomass and wetness regimes and extended the land cover classification for the full Lena Delta with consistent Google Earth Engine aggregated Sentinel-2 reflectance covering the summer 2018 period. We also produced biomass maps derived from Sentinel-2 at a pixel size of 20 m investigating several techniques. The final biomass product for the central Lena Delta shows realistic spatial patterns of biomass distribution, and also showing smaller scale patterns. However, patches of high shrubs in the tundra landscape could not spatially be resolved by all of the landscape-level land cover and biomass remote sensing products.
Biomass is providing the magnitude of the carbon flux, whereas stand age is irreplaceable to provide the cycle rate. We found that high disturbance regimes such as floodplains, valleys, and other areas of thermo-erosion are linked to high and rapid above ground carbon fluxes compared to low disturbance on Yedoma upland tundra and Holocene terraces with decades slower and in magnitude smaller above ground carbon fluxes.

Earth’s population is still growing. The percentage of population living in urban regions was only 30% in 1950, increasing to 55% in 2018. It is projected that 68% of the population will live in urban areas by 2050. As so many people live in urban environments, the topic of urban climate affecting quality of life and public health is of great importance. Several studies found that urban heat stress negatively affects population living in more urbanized regions.

The surface urban heat island (SUHI) effect occurs when an urban area is warmer than its surroundings. Usually, it is computed as the difference in temperature of the rural region relative to the urban core region. The present work uses Land Surface Temperature (LST) data, retrieved through the Meteosat Second Generation geostationary satellite with 3 km at nadir, every 15 minutes.

Paris, Madrid and Milan were chosen as case studies to evaluate how the SUHI varies along the day/year and how does the rural land cover affect the surface heat island intensity. We found diurnal and seasonal variability of SUHI between cities, as a result of different climates.
Results also show that computing the SUHI against different rural land covers results not only in different SUHI intensities but also in different diurnal and seasonal cycles due to the seasonality of the rural land cover. This has consequences when analyzing trends of SUHI for there is much land use and land cover change in the region surrounding cities. This implies that some of the variability and trends in SUHI may be attributed not only to the urban region but also the rural one.

The time dimension is also a key factor, as SUHI intensity and even signal can change throughout the day. Peak intensity may be reached at different times of day/year and poor temporal resolution may not capture/represent this dynamic behavior.

In summary, our results are twofold: (1) noticing the importance of rural land cover as an equal part in the urban/rural relationship in the SUHI topic and (2) stressing that low temporal resolution data, although useful for its spatial characteristics, only tells half the story when considering the SUHI variability.

Climate change has caused dramatic reductions in Earth’s ice cover, which has in turn affected almost all other elements of the environment including global sea level, ocean currents, marine ecosystems, atmospheric circulation, weather patterns, freshwater resources, and the planetary albedo. Here, we combine Earth Observation data and numerical models to quantify global ice losses over the past three decades across the principal components of Earth’s ice system: Arctic sea ice, Southern Ocean sea ice, Antarctic ice shelves, mountain glaciers, the Greenland ice sheet, and the Antarctic ice sheet. Just over half of the ice loss was from the Northern Hemisphere, and the remainder was from the Southern Hemisphere. The rate of ice loss has risen since the 1990s, owing to increased losses from mountain glaciers, Antarctica, Greenland and from Antarctic ice shelves. During this period, the loss of grounded ice from the Antarctic and Greenland ice sheets and mountain glaciers raised the global sea level by more than 3.5 centimetres. The majority of all ice losses were driven by atmospheric melting (from Arctic sea ice, mountain glaciers, ice shelf calving and ice sheet surface mass balance), with the remaining losses (from ice sheet discharge and ice shelf thinning) being driven by oceanic melting. These data improve knowledge of the state of Earth’s cryosphere, a key climate indicator tracked by the EEA and ECMWF, and can be used to help improve the climate models which support decision making in climate mitigation and adaptation. Earth’s ice is also a major energy sink in the climate system; altogether, these elements of the cryosphere have taken up 3 % of the global energy imbalance. Monitoring Earth’s energy imbalance is fundamental in understanding the evolution of climate change and improving climate syntheses and models, and our improved estimates can contribute towards phase 1 of the UNFCCCs global stocktake required by Article 14 of the Paris Agreement, providing information which can be used in testing the effectiveness of climate mitigation policy.

It is well known that the African rainfall climate is highly variable, both in space and time, with many African societies poorly equipped to manage such variability. Access to long-term and regularly updated rainfall information is therefore essential in both drought and flood monitoring and assessment of long-term changes in rainfall. Since gauge records alone are too sparse and inconsistent over time across many parts of Africa, satellite-based records are the only viable alternative, especially in regions with little or no gauges. The longevity of the Meteosat programme, commencing in the late 1970s and running to the present day, thus provides 40 years of continually updated satellite records for monitoring the current climate and assessing long-term changes in rainfall.

Since the early 1980s, the TAMSAT Group (University of Reading) have been providing locally calibrated, operational rainfall estimates based on Meteosat thermal infra-red imagery for Africa. These rainfall estimates are used in a wide range of applications and sectors, as well as in research. While the essence of the TAMSAT estimation algorithm has changed little in four decades, the TAMSAT Group are continually striving to improve the skill and usability of the rainfall products we create.

In this talk, we will present an overview of the TAMSAT rainfall estimation approach as well as a new robust method for combining contemporaneous rain gauge information with the satellite estimates for improving estimation of rainfall amount. A novel feature of this work includes the estimation of spatially coherent rainfall uncertainty – a quantity which is often neglected in operational products but which can greatly support decision making amongst users, especially during adverse weather events. Such developments in TAMSAT have been developed in collaboration with several African organisations to support climate services in regions extremely vulnerable to climate variability and change. We will also highlight capacity building efforts, supported by the World Meteorological Organisation and leading African organisations responsible for issuing agrometeorological advisories, to help facilitate the uptake of TAMSAT products across Africa.

The German Research Centre for Geosciences (GFZ) maintains the “Gravity Information Service” (GravIS, gravis.gfz-potsdam.de) portal in collaboration with the Technische Universität Dresden and the Alfred-Wegener-Institute (AWI). The essential objective of this portal is the dissemination of user-friendly mass variation data in the Earth system based on observations of the German-US American satellite gravimetry missions GRACE (Gravity Recovery and Climate Experiment, 2002-2017) and its successor GRACE-FO (GRACE-Follow-On, since 2018).

The provided data sets comprise products of mass changes of the ice sheets in Greenland and Antarctica, terrestrial water storage (TWS) variations over the continents, and ocean bottom pressure (OBP) variations from which global mean barystatic sea-level rise can be estimated. All data sets are provided as time series of regular grids, as well as in the form of regional basin averages. The ice-mass change is provided either on a regular 50km by 50km stereographic grid or as basin averages which are accompanied by realistic uncertainties. The gridded continental TWS data, as well as the OBP data, are given on a 1° by 1° grid. For continental TWS data, the user can choose between river discharge basins and segmentation based on climatically similar regions. All regional mean time series of the TWS product are accompanied by realistic uncertainty estimates. The OBP data set is composed of a barystatic sea-level map and a map of the residual ocean circulation which was not reduced by background models in the data processing. These background models are also provided for all three data products.

The data sets of all domains can be interactively displayed at the portal and are freely available for download. This contribution aims to show the features and possibilities of the GravIS portal to researchers without a dedicated geodetic background in the fields of climatology, hydrology, cryosphere, or oceanography. The data provided on the portal will also be used within the GRACE-FO project of the ESA Third Party Mission Program.

The International Soil Moisture Network (ISMN, https://ismn.earth) is a unique centralized global and open freely available in-situ soil moisture data hosting facility (Dorigo et al.,2021: https://hess.copernicus.org/articles/25/5749/2021/). Initiated in 2009 as a community effort through international cooperation (ESA, GEWEX, GTN-H, WMO, etc.), the ISMN is more than ever an essential means for validating and improving global satellite soil moisture products, land surface -, climate- , and hydrological models.

Following, building and improving standardized measurement protocols and quality techniques, the network evolved into a widely used, reliable and consistent in-situ data source (surface and sub-surface) collected by a myriad off data organizations on a voluntary basis. 72 networks are participating (status November 2021) with more than 2800 stations distributed on a global scale and a steadily increasing number of user community, ~ 4000 registered users strong. Time series with hourly timestamps from 1952 – up to near real time are stored in the database and are available through the ISMN web portal for free (https://ismn.earth), including daily near-real time updates from 7 networks (~ 1000 stations).

More than 10’000 in-situ soil moisture datasets are available through the web portal and the number of networks and stations covered by the ISMN is still growing as well as most datasets, that are already contained in the database, are continuously being updated.
The ISMN evolved in the past decade into a platform of benchmark data for several operational services such as ESA CCI Soil Moisture, the Copernicus Climate Change (C3S), the Copernicus Global Land Service (CGLS), the online validation service Quality Assurance for Soil Moisture (QA4SM) and many more applications, services, products and tools. In general, ISMN data is widely used in a variety of scientific fields with hundreds of studies making use of ISMN data (e.g. climate, water, agriculture, disasters, ecosystems, weather, biodiversity, etc.).

The foundation and continuous development of the ISMN was funded by the European Space Agency (formerly SMOS and IDEAS+ programs, currently QA4EO program). However, it was always clear that financial support from ESA was not realizable on a long term basis. Therefore, several different options for financing ISMN where explored within the last couple of years together with ESA.
In January 2021, the German Federal Ministry of Transport and Digital Infrastructure (BMVI: https://www.bmvi.de/EN/Home/home.html) agreed to provide continuous long-term funding for the ISMN operations. Three full time positions are financed at the German Federal Institute for Hydrology (Bafg: https://www.bafg.de/EN/) as well as two full time positions at the connected International Center for Water Resources and Climate Change (ICWRGC https://www.waterandchange.org/en/ - under the auspice of UNESCO and WMO). The transfer of the ISMN operations from Austria (TU Wien) to Germany started in May 2021 and will be finished by end of 2022. This 19 month transfer timeframe is co-financed by ESA and the German Ministry to facilitate a sustainable transfer of knowledge and operations.

In this session, we want to introduce the new hosts (BafG and ICWRGC) and look back at the evolution of the ISMN over the past decade (network and dataset updates, quality procedures, literature overview, and current limitations in data availability – functionality and challenges in data usage). Furthermore, we want to especially look ahead and share new possibilities for the ISMN to serve the EO community for a long time to come.

Climate change indicators are designed to support climate policy making and public discussions. They are important for setting, monitoring and evaluating targets and communicating changes of the investigated phenomenon. Impact indicators highlight how climate change affects certain environmental phenomena. Response indicators show how society adapts to climate change. In Germany, the Environmental Federal Agency (Umweltbundesamt) coordinates the German Adaptation Strategy to climate change. This framework comprises around 100 impact and response indicators in six clusters, i.e., health, water, land, infrastructure, economy and spatial planning/ civil protection. Indicator assessment on a national scale demands comparable data of national scale ; comparability but even availability of environmental data, however, is often challenging. Lakes, for instance, are considered as sentinels of climate change, but nation-wide data for consistent and long time series are rare.
Remote sensing of lakes experiences significant developments during the last decade. Thus, the next report of the German Adaptation Strategy aims to include remote sensing data and methods for the first time. The focus lies on four impact indicators in lakes, namely “presence of cyanobacteria” (cluster health), “beginning of spring phytoplankton bloom”, “lake water temperature” and “ice cover” (cluster water). The aim of our project is to develop an operational, retrospective processing routine based on remote sensing data for these four climate change indicators. We collected a large in-situ database for 25 lakes in Germany, for which we tested and evaluated potentially suited algorithms and sensors. We also discussed with experts and end-users the requirements on sensor-algorithms. Then, we developed different approaches to create and visualise the indicators, i.e., to obtain an easily –to-grasp figure from the remote sensing data. The results are briefly summarised below:
“Presence of cyanobacteria”:
ENVISAT MERIS and Sentinel-3 OLCI data form the data basis, Sentinel-2 is in preparation. The maximum Peak Height algorithm is used to determine presence or absence of cyanobacteria. To aggregate at lake level, we count the days with cyanobacteria presence during the season (March to October) and summer (June to September). As basis for the indicator, we set the number of days with cyanobacteria presence in relation to the number of valid image acquisitions.
“Beginning of spring phytoplankton bloom”:
ENVISAT MERIS, Sentinel-3 OLCI and Sentinel-2 MSI data form the data basis. We calculate chlorophyll-a concentrations using C2X-COMPLEX (Sentinel-2 MSI), merged algorithm derived from Maximum Peak Height following Pitarch calibration (Sentinel-3 OLCI) and C2RCC (ENVISAT MERIS) of all suited imagery acquired from March to May. The percentile 90 is used to aggregate at lake level to detect spatially variable spring blooms. From the time series, we extract the day of year and week of year at which chlorophyll-a concentration peak exceeds the 70 percentile for the first time during spring. This date then is considered as beginning of spring bloom.
“Lake water temperature”:
The Landsat 5 TM, 7 ETM and 8 TIRS thermal data form the data basis. We selected the mono-window algorithm by Sobrino/ Jimenez-Munoz combined with ERA5-Land data to retrieve lake surface water temperature. Investigation on Landsat-8 collection-2 performance is ongoing. The subsequent data analysis homogenises the results to Landsat 8 and filters outliers. The median is used to aggregate to lake level, which are then temporally averaged to monthly data. We interpolate missing monthly data if gaps are not exceeding 1 month. Gaps occur over all the year due to low revisit time of Landsat and cloud coverage. Yearly seasonal (March to October) and summer averages (June to August) are the basis for the indicator.
“Ice cover”:
Landsat 8 OLI, Sentinel-2 MSI and Sentinel-1 data form the data basis. We developed sensor-specific random forest classification models to separate ice and water and mask out clouds (only optical imagery). To aggregate to lake level, we determine days when ice covers more than 80 % of the lake. Then we count the number of ice days and calculate the ratio of ice days and the number of valid image acquisitions.
Based on the above-mentioned approaches and discussions with stakeholders, we developed a framework to evaluate the data quality for the indicators. This framework indicates spatial and temporal measures of data coverage for assessing the representativeness of a value to be included in the long-term trends. Such quality measures support calculating reliable trends. Currently, we transfer the developed approaches into a retrospective, operational service for the German Environmental Agency using the cloud-processing structure of CODE-DE (National Collaborative Ground Segment). In a next step, we calculate trends and examine whether similar patterns can be derived among groups of lakes or on a national level.
Our presentation will focus on the transfer of pixel-based information into a climate change indicator, the experienced challenges, but also the new opportunities.

An accurate monitoring of the snow-albedo feedback is essential for understanding the effects of climate change in snow-covered regions. The IPCC’s Sixth Assessment Report (AR6) established that a surface albedo feedback in the range of +0.35 [0.10 to 0.60] Wm-2C-1 is very likely [1]. The main component of this feedback is the so-called snow/ice-albedo feedback, which until AR5 was analyzed independently. AR6 included as well temperature-induced albedo changes over snow-free surfaces. Snow/ice-albedo feedback has been generally monitored with global climate models (GCMs). The increasing availability of satellite observations provides new opportunities to reduce the uncertainty in the snow-albedo feedback estimates, and also to improve its understanding by separating the contribution of ice and snow, and within snow, by separating the contribution of snow cover retreat and snow metamorphosis [2]. Indeed, observations are being increasingly used either to constrain GCMs [1], or to estimate the snow albedo feedback directly from multi-decadal observations [3].

Two types of observational products are currently being used: satellite-based products and global reanalyses. However, both face stability challenges that need to be quantified to understand the uncertainty of the snow-albedo feedback estimates obtained. Satellite products concatenate different sensors (e.g., C3S albedo) or different versions of the same sensor (e.g., AVHRR, VGT), which can introduce discontinuities during the transition periods. For each sensor, orbital drifts and instrument degradation are also a problem. Additional instabilities are added by the retrieval algorithm and the snow mask used. Besides, the uncertainty of albedo retrievals increases over snow due to the highly anisotropic reflectance of snow and the generally low solar angles during snow albedo retrievals.

Stability issues in reanalysis are related to the addition of new observations (satellite or ground) into the data assimilation system. Reanalyses face a trade-off between accuracy and stability that depends on the weight they give to new observations. NWP initialization applications require more accurate estimations that are obtained by weighting more recent observations, which generally introduces temporal instabilities in the long-term. By contrast, climate applications prefer stability over accuracy. Therefore, instabilities of different degrees can be present in reanalysis products based on the approach undertaken.

Our goal is to evaluate if the existing satellite and reanalysis products are fit for monitoring the snow-albedo feedback. The satellite products evaluated are MCD43C3 v6.1 (2000-present), CLARA-A2.1 (1982-present), GLAS-AVHRR v4 (1982-present), and C3S v2 (1982-present). The reanalyses evaluated are ERA5 (1950-present), ERA5-Land (1950-present), MERRA-2 (1982-present), and JRA-55 (1958-present). First, we evaluate if snow albedo values and trends from the different products are consistent globally. Then, we quantify how instabilities and incostinencies in multi-decadal albedo datasets propagate to the snow-albedo feedback estimates. For that, we generate an independent estimate of the snow-albedo feedback from each product using a common radiative kernel [4]. Our final aim is to determine whether the existing products area accurate and stable enough, and to identify aspects that can be improved to reduce the uncertainty of snow-albedo feedback estimates.

Bibliography

[1] IPCC, Climate Change 2021: The Physical Science Basis Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change [Masson-Delmotte V, Zhai P, Pirani AS, Connors SL, Péan C, Berger S, Caud N, Chen Y, Goldfarb L, Gomis MI, Huang M, Leitzell K, Lonnoy E, Matthews JBR, Maycock TK, Waterfield T, Yelekçi O, Yu R, Zhou B (eds.)]. Cambridge University Press. In Press.]
[2] Wegmann M, Dutra E, Jacobi HW, Zolina O. Spring snow albedo feedback over northern Eurasia: Comparing in situ measurements with reanalysis products. The Cryosphere 12, 1887-1898, 2018
[3] Xiao L, Che T, Chen L, Xie H, Dai L. Quantifying snow albedo radiative forcing and its feedback during 2003–2016. Remote Sensing 9, 883, 2017
[4] Pithan F, Mauritsen T. Arctic amplification dominated by temperature feedbacks in contemporary climate models. Nature Geoscience 7, 181-184, 2014.

Cities are warmer than their surroundings. This phenomenon is known as Urban Heat Island (UHI) and is one of the clearest examples of human-induced climate modification. Surface UHIs (SUHI) result from modifications of the surface energy balance at urban facets, canyons, and neighborhoods. The difference between urban and rural Land Surface Temperatures (LST)—known as SUHI Intensity (SUHII)—varies rapidly in space and time as the surface conditions, the weather, and the incoming radiation change, and is generally strongest during daytime and summertime. In this work we revisit the topic of SUHII seasonality and how it differs across climates. Our thesis is that aggregating global SUHII data without considering the biome (i.e., vegetation zone) of each city, can lead to erroneous conclusions and estimates that fail to reflect the actual SUHII characteristics. This is because SUHII is a function of both urban and rural features, and the phenology of the rural surroundings can differ considerably between cities even in the same climate zone. To test this hypothesis, we use 18 years (2000-2018) of global land cover and MODIS LST data from the European Space Agency’s Climate Change Initiative (ESA-CCI). Our analysis covers 1588 cities in 12 tropical, dry, temperate, and continental Köppen-Geiger sub-classes. This classification scheme empirically maps Earth in 5 main and 30 sub- classes by assuming that vegetation zones reflect climatic boundaries. To analyze our results, we calculate, for each climate class, the seasonal variation of SUHII and rural LST (at monthly resolution) by averaging the corresponding city data (we do this separately for daytime and nighttime). Our results reveal that the seasonality of tropical, dry, temperate, and continental SUHIs differs considerably during daytime and that it is more pronounced in temperate and continental climates. They also show that the seasonality of the dry and temperate sub-classes exhibits considerable intra-class variation. In particular, the month when the daytime SUHII is strongest can differ between temperate sub-classes by as much as 4 months (e.g., for the hot-Mediterranean sub-class it occurs in May and for the dry-winter subtropical highlands sub-class in September), while the corresponding SUHII magnitude by as much as 2.5 K. The strong intra-class variation of temperate climates is also evident in the corresponding hysteresis loops, where almost every sub-class exhibits a unique looping pattern. These finding support our thesis, and suggest that global SUHII investigations should consider, in addition to climate, and the distribution of biomes when aggregating their results. Our results provide the most complete typology of SUHII hysteresis loops to date and an in-depth description of how SUHIIs vary within the year across climates.

The present work shows the potential of satellite thermal observations to estimate Earth’s global surface temperature trends and, therefore, its applicability to climate change studies. Present satellites allow estimation of surface temperature for a full coverage of our planet with a sub-daily revisit frequency and kilometric resolution. In this work, a simple methodology is presented that allows estimating the surface temperature of Planet Earth with MODIS Terra and Aqua land and sea surface temperature products, as if the whole planet was reduced to a single pixel. The results corroborate the temperature anomalies retrieved from climate models and show a rate of warming higher that 0.2 °C per decade. In addition Earth’s surface temperature are analysed in more detail over the period 2003-2021 by dividing the globe into the northern (HN) and southern (HS) hemispheres, and each hemisphere into three additional zones: the low latitudes from the Equator to the Tropic of Cancer in the HN and Tropic of Capricorn in the HS (0-23.5⁰), mid latitudes from the Tropics to the Arctic Circle in the HN and Antarctic Circle in the HS (23.5⁰-66.5⁰) and high latitudes from the Arctic and Antarctic Circles to the Poles (66.5⁰-90⁰).

Lake ice cover (LIC), a thematic variable under Lakes as an Essential Climate Variable (ECV) that is a robust indicator of climate change and plays an important role in lake-atmosphere interactions at northern latitudes (i.e. heat, moisture, and gas exchanges), refers to the area (or extent) of a lake covered by ice. Ice dates and ice cover duration at the pixel scale (ice-on and ice-off) and lake-wide scale (complete freeze-over (CFO) and water clear of ice (WCI)) can be derived from lake ice cover data (Duguay et al. 2015). Determination of ice onset (date of the first pixel covered by ice), CFO, melt onset (date of the first pixel with open water), and WCI are of most relevance to capture important ice events during the freeze-up and break-up periods. Duration of freeze-up and break-up periods and duration of ice cover over a full ice season can be determined from these dates. The generation of a LIC product from satellite observations requires the implementation of a retrieval algorithm that can correctly label pixels as either ice (snow-free and snow-covered), open water, or cloud. The LIC product v2.0 generated for Lakes_cci (https://climate.esa.int/en/projects/lakes/) uses MODIS Terra/Aqua data to provide the most consistent and longest daily historical record globally to date (2000-2020). The new product provides three bands: Band 1 - lake ice cover flag (lake forms or does not form ice); Band 2 - lake ice cover class (open water, ice, cloud, and bad); and Band 3 - lake ice cover uncertainty (% accuracy for each of open water, ice and cloud classes).
In the first step of production, the Canadian Lake Ice Model (CLIMo) was applied to help determine which lakes of the Lakes_cci harmonized product (total 2024 lakes), which includes four other variables (water level, water extent, surface water temperature, and water-leaving reflectance), could have formed ice or have remained ice-free at any time over the 2000-2020 period. This step can correct false detection of ice in summer in the situation of dry lakebeds and reduce the computational cost of the production. CLIMo (Duguay et al. 2003) is a one-dimensional thermodynamic model capable of simulating ice phenology events, ice thickness and temperature, and all components of the energy/radiation balance equations during the ice and open water seasons at a daily timestep. Input data to drive CLIMo include mean daily air temperature (°C), wind speed (m s-1), relative humidity (%), snowfall (or depth) (m), and cloud cover (in tenth). Here, European Centre for Medium-Range Weather Forecasts (ECMWF) ERA5 reanalysis hourly data on single levels (0.25-degree grid) were used to generate inputs required for CLIMo simulations for each of the 2024 lakes. Lake ice depth data provided by ERA5 were also utilised to check for the possible formation of ice on any of the lakes. Ice cover was deemed possible to have formed on a lake if ice depth was determined to have reached a thickness greater than 0.001 m on any day from either CLIMo or ERA5. Additionally, as a third check, a number of lakes (largely located at the southern limit of where ice could potentially form during a cold winter in the Northern Hemisphere and in mountainous regions of both the Northern and Southern hemispheres) were inspected manually through interpretation of MODIS RGB images to determine if any of these lakes had formed ice between 2000 and 2020. As a result of the process described above, presented in the variable of lake ice cover flag of the LIC product v2.0, 1391 of 2024 lakes were flagged as forming an ice cover and 633 not forming any ice over the 2000-2020 period. Once flagged, only lakes determined to form ice were selected to perform lake ice classification from MODIS data by the main processing chain.
MODIS TOA reflectance bands and the solar zenith angle (SZA) band are used for feature retrieval (i. e. for labeling as water, ice, or cloud) (Wu et al. 2021). The reflectance bands are MOD02QKM at 250 m (band 1: 0.645 µm and band 2: 0.858 µm) and MOD02HKM at 500 m (band 3: 0.469 µm; band 4: 0.555 µm; band 5: 1.240 µm; band 6: 1.640 µm; band 7: 2.130 µm) resolutions. Prior to retrieval, pixels of interest are identified as “good” or “bad” using quality bands from the original MODIS TOA reflectance product. The pixels with SZA greater than 85 degrees are identified as “bad”. Pixels of interest are classified and labelled as either cloud, ice, or water from a random forest algorithm (Wu et al. 2021). Labelled pixels are resampled to the output grid. The processing chain has been revised for Lakes_cci to generate the output grid based on specifications of the harmonized product (1/120th degree latitude/longitude; ca. 1 km). Aggregation is performed by taking a majority vote between ice and water, ties broken by selecting water. If there are zero ice and water pixels, then the cell is labelled as cloud if there are non-zero cloud pixels; otherwise, the output cell is labelled as “bad”. The variable of lake ice cover class presents the retrieved labels.
Validation of the LIC V2.0 product has been performed through the computation of confusion matrices built on independent statistical validation. The reference data for validation were collected for water, ice, and cloud as AOIs from the visual interpretation of the MOD02/MYD02 false color composite images (R: band 2, G: band 2, B: band 1) with a 250 m spatial resolution. A total of 10,075,081 pixels taken from 229 MOD02 swaths over Great Slave Lake and Lake Onega were used to conduct classification assessment of the LIC product generated by MODIS Terra. There is no notable difference in the accuracy of the product between the break-up (98.14% overall accuracy) and freeze-up (96.83% overall accuracy) period. Additionally, 1,665,188 samples collected from MYD02 false color composite images were applied for the validation of the LIC product produced from MODIS Aqua. The overall accuracy of 97.68% reached with Aqua data is comparable to that obtained with MODIS Terra data. Further evaluation of the Lakes_cci LIC V2.0 product and its comparison with other products is planned in the future, and with input from the user community.

References
Duguay, C. R. Bernier, M. Gauthier, Y. & Kouraev, A. (2015). Remote sensing of lake and river ice. In Remote Sensing of the Cryosphere, Edited by M. Tedesco. Wiley-Blackwell (Oxford, UK), 273-306.
Duguay, C.R., Flato, G.M., Jeffries, M.O., Ménard, P., Morris, K. & Rouse, W.R. (2003). Ice cover variability on shallow lakes at high latitudes: Model simulations and observations. Hydrological Processes, 17(17), 3465-3483.
Wu, Y., Duguay, C.R. & Xu, L. (2021). Assessment of machine learning classifiers for global lake ice cover mapping from MODIS TOA reflectance data. Remote Sensing of Environment, 253, 112206, https://doi.org/10.1016/j.rse.2020.112206.

The Greenland Ice Sheet has had a negative mass balance over at least the last two decades, during which there has been a well documented increase in the retreat of the ice sheet. Increased dynamic thinning and lower surface mass balance are roughly equally important mechanisms behind the continuous reduction of the Greenland Ice Sheet, with the latter largely being driven by enhanced melt and run-off rates. In the continuous effort to better simulate the evolution of the Greenland Ice Sheet under different climate change scenarios, models calculate the surface energy budget and convert this to ice surface temperature (IST) in order to calculate melt and run-off. Accurately characterising the ice surface temperature is essential, as it regulates surface melt and run-off through various mechanisms.

Surface temperature monitoring over the polar regions is impeded by harsh environmental conditions, making in situ monitoring challenging and scarce. Space-borne retrievals of ice surface temperature are challenging due to complications from persistent cloud cover, large daily temperature variations and the lack of high quality in-situ observations for validation. Nonetheless, a continuous effort calibrating and harmonising the extended archive of surface temperatures from various sensors has now resulted in comprehensive IST datasets spanning over nearly four decades. A significant part of these datasets are available in satellite processing levels L2 (swath) and L3 (gridded on regular grid) yet with gaps due to cloud cover. Optimally interpolated products offer gap-free fields, typically on a daily basis and while there is a suite of global coverage datasets, few have specifically been developed for the Arctic region.

This study reports from a user case study (UCS) conducted within the ESA CCI LST project. The aim of the UCS was to use the L2 ESA CCI LST products along with the L2 Arctic and Antarctic Ice Surface Temperatures from thermal Infrared satellite sensors (AASTI) v2 dataset, to develop a L4 optimally interpolated, multi-sensor, gap-free, surface temperature field for the Greenland Ice Sheet. The L4 product was produced daily for the year 2012 with a spatial resolution of 0.01 degree latitude and 0.02 degree longitude. Prior to the generation of the gap-free daily fields, the upstream input data were inter-compared and a cold bias for LST CCI MODIS retrievals was identified and corrected against the AASTI dataset. All L2 input data along with the derived product were validated using observations from the PROMICE automated weather stations (AWS) on the Greenland Ice Sheet as well as the IceBridge flight campaigns. L2 AASTI and the L4 OI field shared similar bias and standard deviation values, while MODIS demonstrated a cold bias. The L4 OI fields were used to examine the monthly and seasonal variability of IST during 2012 when a significant melt event occurred. Mean surface temperature for July was around zero for the largest part of the Greenland Ice Sheet, based on the aggregation of 200 to 700 observations depending on the region. Melt days, defined as days when IST was -1°C or higher, ranged between 5 and 10 for the central part of the Greenland Ice Sheet and exceeded 30 for the middle and lower zones in the periphery of the ice sheet. The L4 OI product was assimilated into a surface mass balance (SMB) model of the Greenland Ice Sheet to examine the impact of the multi-sensor, gap free dataset on modelled snowpack properties that account for important effects including refreezing and retention of liquid water for the test year of 2012.

The surface soil moisture (SM) state impacts the sphere of human-nature interaction at different levels. It contributes to changing the frequency and extent of extreme atmospheric events such as heatwaves, and it affects the ecosystem state which anthropogenic activities depend on. SM is therefore recognized as an Essential Climate Variable (ECV). Its monitoring at scales from multi-decadal to near-real time (NRT) benefits study fields as diverse as agricultural crop yield forecasting, wildfires prediction or drought and flood risk management.

The European Commission’s Copernicus Climate Change Service (C3S) includes a soil moisture data set that is regularly updated to support timely decision making. The C3S SM product is freely made available through the Copernicus Climate Data Store with a global coverage at daily, 10-daily and monthly aggregation levels. It integrates multiple NRT data streams for this purpose: The Land Parameter Retrieval Model (LPRM; Owe et al., 2001) is used to derive SM from operational satellite radiometers (AMSR2, SMAP, SMOS and GPM). EumetSat HSAF produces a scatterometer based SSM product from ASCAT sensors (on board of Metop-A/B/C) with a short delay (HSAF, 2019). Using a modified version of the ESA CCI SM merging algorithm (Gruber et al., 2019), C3S SM can therefore provide an ACTIVE (scatterometric), PASSIVE (radiometric) and COMBINED product with a short delay of 10-20 days. The C3S SM algorithm is updated on an annual basis with the latest scientific improvements from ESA CCI SM. Products are validated with in-situ measurements from the International Soil Moisture Network (ISMN; Dorigo et al., 2021) and reanalysis reference data using the QA4SM online validation service. Assessment reports are distributed with the data sets.

Several derived services can greatly profit from the use of C3S SM due to its short update delay. One outstanding example is the detection and estimation of precipitation amounts at the regional scale as performed in the SM2RAIN project (led by Italy’s IRPI-CNR institute), with applications in drought and flood analysis and management. Similarly, the impact of climatic extremes on food security can be mitigated using the scientific knowledge basis provided by C3S SM, as demonstrated in the EarthFoodSecurity service. This presentation will cover the climate service provided with C3S SM, including the input data streams, the processing and distribution of the products and their quality assessment; the impact and external applications of the service will also be covered.

The development of the ESA CCI products has been supported by ESA’s Climate Change Initiative for Soil Moisture (Contract No. 4000104814/11/I-NB and 4000112226/14/I-NB) and the European Union’s FP7 EartH2Observe “Global Earth Observation for Integrated Water Resource Assessment” project (grant agreement number 331 603608). Funded by Copernicus Climate Change Service implemented by ECMWF through C3S 312a/b Lot 7/4 Soil Moisture service.


References

Dorigo, W., Himmelbauer, I., Aberer, D., Schremmer, L., Petrakovic, I., Zappa, L., ... & Sabia, R. (2021). The International Soil Moisture Network: serving Earth system science for over a decade, Hydrol. Earth Syst. Sci., 25, 5749–5804, https://doi.org/10.5194/hess-25-5749-2021, 2021.

Gruber, A., Scanlon, T., van der Schalie, R., Wagner, W., & Dorigo, W. (2019). Evolution of the ESA CCI Soil Moisture climate data records and their underlying merging methodology. Earth System Science Data, 11(2), 717-739.

H-SAF (2019) ASCAT Surface Soil Moisture Climate Data Record v5 12.5 km sampling - Metop (H115), EUMETSAT SAF on Support to Operational Hydrology and Water Management, DOI: 10.15770/EUM_SAF_H_0006.

Owe, M., de Jeu, R., & Walker, J. (2001). A methodology for surface soil moisture and vegetation optical depth retrieval using the microwave polarization difference index. IEEE Transactions on Geoscience and Remote Sensing, 39(8), 1643-1654.

The Microwave Radiometer (MWR) represents a series of nadir viewing instruments whose main purpose is to provide the information required to correct ocean altimeter observations for the highly variable effects of atmospheric water vapour (the ‘wet tropospheric correction (WTC)’). MWR instruments have been flown onboard the ERS-1 (1991-2000), ERS-2 (1995-2011), and Envisat (2002-2012) platforms and are recently flown again onboard the Sentinel-3 series of satellites (S3-A 2016 - ongoing, S3-B 2018 - ongoing).

The MWR instrument also allows for an accurate determination of the atmospheric total column water vapour (TCWV), under clear and cloudy sky conditions, during both day and night.

In our presentation, we report on recent activities to derive a consistent high-quality long-term TCWV and WTC dataset from MWR observations. A novel bias correction method is applied to create bias-free cross-instrument brightness temperature time series as well as the corresponding TCWV values using a 1D-VAR approach.

The aim of these activities is to create a TCWV and WTC data record that covers the entire 30+ year period from 1991 to 2021 (except for the four year data gap between Envisat and S3-A).

Aside from its immediate contribution to altimetry, MWR-derived TCWV retrievals have the potential to play an important role in climatology and the validation of other TCWV retrievals.

The Copernicus Atmosphere Monitoring and Climate Change Services (CAMS and C3S respectively), two of the six core Services of the Copernicus programme, enter an exciting new phase with the signature of a Contribution Agreement between ECMWF and the European Commission in July 2021.
Both Services are fully operational and routinely deliver a wide variety of environmental products, based on Sentinel and other satellite data, in-situ observations and modelling information. These data and products are accessed by hundreds of thousands of users.
A unique and strong point of ECMWF Copernicus Services is to focus on delivering operationally “authoritative” data, via their Copernicus data stores. A unique and strong point of C3S is to focus on delivering operationally “authoritative” data, via its climate data store. C3S provides authoritative information about the past, present and future climate, as well as tools to enable climate change mitigation and adaptation strategies by policy makers and businesses, while CAMS delivers consistent and quality-controlled information related to air pollution and health, solar energy, greenhouse gases and climate forcing, everywhere in the world. A prime user of both Services is the European Commission itself, and CAMS and C3S strive to support the policy makers and public authorities by providing the environmental information they need to inform their policies and legislation, etc. which becomes critically important in view of following up the Paris Agreement, and supporting the UNFCCC Sustainable Development Goals, the Sendai Framework for Disaster Risk Reduction and the Green Deal, to name a few.
This presentation will provide an overview of the State-of-Play of both Services and their foreseen evolution over the next seven years. We will particularly emphasize the development of the new anthropogenic CO2 emissions Monitoring and Verification Support Capacity (CO2MVS) which will combine satellite observations with modelling information to enable users to precisely pinpoint which component of emissions are resulting from human activity.

Today, information to support carbon emission control and carbon assimilation by forests is of very variable quality. The information sources are versatile: different types of field data, aerial photography, laser scanning data, and satellite imagery. This information is used as input for calculation models using which it is decided whether forest is a carbon sink or an emission source. The results can be further used to value the carbon on the growing market of the voluntary carbon trade.

In future, it will be ever more important that forest owners, governments, academia, investors, and organizers of the voluntary carbon market and can base their decisions on as accurate, reliable and comparable information as possible and that the information is easily accessible.

In the VTT-led Horizon 2020 Innovation Action project Forest Flux a service was developed to offer reliable and comparable information on forest resources and forest carbon. Forest Flux cloud service on the Forestry Thematic Exploitation Platform F-TEP includes a seamless service chain from field observations and satellite imagery. It produces estimates of the present forest resources and carbon assimilation on a certain area and their future forecasts. The forecasts can be computed by applying different climate scenarios. It is to our knowledge the first of its kind globally.

The main satellite data source was Sentinel-2 of the Copernicus program. Additional data sources included very high-resolution optical imagery and airborne laser scanning (ALS) data. Ground reference data were provided by the users or were acquired from open sources.

The services were offered for nine users in Finland, Germany, Portugal, Romania, Paraguay, and Madagascar located in the boreal, temperate, and tropical vegetation zones. The user types included private and governmental large forest owners and managers, forest industries, associations of forest owners, and a development aid organization.

The users could select their desired map products from the portfolio of 51 alternatives. They included natural and color infrared image maps, forest cover map, nine traditional forest structural variables, site fertility type, three change map types, five forest fragmentation variables plus five variables for their changes, four biomass variables, nine carbon flux variables plus nine variables indicating their change, and eight variables to forecast the biomass and carbon assimilation. In addition, statistical information on the carbon balance of an organization was computed. Inputs for the organizational carbon balance were, in addition to the satellite image based carbon assimilation products, emissions from the silvicultural measures, harvest, and transportation, provided by the user.

The main method for satellite image analysis was the in-house probability software whose benefit is its adaptivity to different quality and amount of reference data because the models can be checked and modified manually (Häme et al., 2001, 2013). For the mapping of change, another in-house tool Autochange was used (Häme et al., 2020).

The process model PREBAS was used to compute and forecast the primary production variables. It used as inputs the outputs of the structural variable estimation and daily data on temperature and precipitation (Minunno et al., 2019; Tian et al., 2020). The model that was originally developed for boreal forest was parametrized for several other species that grew in the study sites. Comparison of the model predictions with flux tower measurements indicated voi very good match.

Software components for the Forest Flux services we developed for the F-TEP platform where they are applicable for operational services. The processing chain is largely automated. The main challenges were the variable quality, amount, and formats of the reference data as well as residual clouds in the pre-processed imagery, which led to manual work in the development of the models for the estimation of the structural variables. The uncertainty of the results was computed using a random sample from the reference data. However, in some cases, the reference data were not adequate for an independent set for uncertainty assessment and the results had to be assessed using the training data.

The relative root mean square error (RMSE) for the growing stock volume estimation varied between 29% and 67%. The error was always smaller for the other estimated structural variables stem basal area, mean height, and stem diameter than for volume. The bias was usually few percent with an exception at two sites in the same country where the overestimation was over 20% computed with a limited reference data. In Finland, the pure Sentinel-2 based estimation provided a relative error of 45%. By including the ALS data in the model, the RMSE dropped to 31%.

In total, about 1300 raster maps at ten-meter pixel size or vector outputs were computed in two phases. The user feedback was collected after both phases. In the short term, the most desired services concern forest change and the traditional structural variables and biomass. Carbon market is still poorly developed but this is expected to grow fastest within coming few years due to international regulations, and pressure from company shareholders and public.

The three-year Innovation Action project Forest Flux started in 2019 and it was completed in November 2021. The operational services can be started immediately after the completion of the project.

Project partners, in addition to VTT Technical Research Centre of Finland Ltd. were Unique Land Use GmbH (DE), Simosol Oy (FI), University of Helsinki (FI), Instituto Superior De Agronomia (PT), and The National Institute for Research and Development in Forestry (RO). The project was supported by the Horizon2020 Program of the EU, Grant Agreement #821860.

https://www.forestflux.eu/
https://f-tep.com/

Häme, T. et al. (2001) ‘AVHRR-based forest proportion map of the Pan-European area’, Remote Sensing of Environment, 77(1), pp. 76–91. doi: 10.1016/S0034-4257(01)00195-X.

Häme, T. et al. (2013) ‘Improved mapping of tropical forests with optical and sar imagery, part i: Forest cover and accuracy assessment using multi-resolution data’, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 6(1), pp. 74–91. doi: 10.1109/JSTARS.2013.2241019.

Häme, T. et al. (2020) ‘A Hierarchical Clustering Method for Land Cover Change Detection and Identification’, Remote Sensing. MDPI AG, 12(11), p. 1751. doi: 10.3390/rs12111751.

Minunno, F. et al. (2019) ‘Bayesian calibration of a carbon balance model PREBAS using data from permanent growth experiments and national forest inventory’, Forest Ecology and Management. Elsevier B.V., 440, pp. 208–257. doi: 10.1016/j.foreco.2019.02.041.

Tian, X. et al. (2020) ‘Extending the range of applicability of the semi‐empirical ecosystem flux model PRELES for varying forest types and climate’, Global Change Biology. Blackwell Publishing Ltd, 26(5), pp. 2923–2943. doi: 10.1111/gcb.14992.

Flooding is recognised as an environmental hazard that affects more people than any other environmental hazard. It is also anticipated to affect a higher proportion of the global population and incur rising costs in the future due to rapid urbanization, increasing settlements in floodplains, climate change, and variability.

To meet these challenges Previsico have developed their FloodMap Live software to provide high resolution, real-time flood forecasts based on a predictive flood modelling system. Flood forecasts, such as those provided by Previsico, enable actions to be taken to reduce loss of life and property in the event of a flood and help to identify genuine insurance claims post-flood.

Yet flood models require integration and validation with external data sources, such as satellite imagery, for re-calibration of model predictions and to demonstrate prediction effectiveness. Independent information from satellite data enables refinements to be made to flood models, in turn supporting more accurate forecasts of flooding evolution.

Following a successful collaboration with the University of Leicester, Previsico is developing a flood extent product derived from Sentinel 1 radar imagery that will provide near-real-time information on flood location and extent in both urban and rural areas. Synthetic Aperture Radar (SAR) was chosen for the satellite product as data collection is not impeded by cloud cover or a lack of illumination and can acquire data over a site during day or night time under almost all weather conditions. Furthermore the Sentinel 1 SAR-C instrument provides dual polarisation capability, very short revisit times and rapid product delivery. This satellite product will allow Previsico to refine it’s model in order to offer more accurate and validated flood models to their customers ensuring they can respond to a flood event in a targeted and efficient manner.

Here we will present our progress so far in developing and utilising Sentinel 1 SAR data to refine and validate a commercial flood model. Results from the initial version of this Sentinel 1 flood product were encouraging, as shown in Figure 1 for an area over Doncaster and Rotherham in the UK which were affected by flooding in November 2019. The method also performed well in non-flood events suggesting it is fairly robust even when inundation has not occurred. Further comparisons against external data sources such as Copernicus EMS showed promise and allowed us to identify improvements to the code, either to be implemented in the prototype product or in future versions of this product. Comparisons to flood forecasts from Previsico’s flood modelling system were also performed and the results from this will be presented.

This presentation will present the plans of EUMETSAT's Network of Satellite Application Facilities (SAFs) for the period 2022 - 2027. The SAF Network consists of eight Satellite Application Facilities dedicated to provide operational services for specific application areas. One element is the sustained generation of Climate Data Records from satellite data to support climate science and climate services. In 2021 the commitments for a fourth "Continuous Development and Operations Phase (CDOP4) have been approved. An overview of the Climate Data Record portfolio, the applied concepts as well as the applications examples will be presented.

With the rising awareness and visibility of impacts caused by climate change and linked extreme weather events, the need for rapid dissemination of and access to information is becoming a progressively pressing matter in many different anthropogenic, social and economic sectors. To meet these needs, the existing wealth of free, open and globally available analysis-ready weather and climate data serves as a valuable source.

However, the lack of understanding how to access, handle and combine data sets from different sources often prevents end-users in the previously mentioned sectors from making use of the data. Furthermore, domain-specific knowledge to extract additional information out of climate and weather data is hardly existing.

Because of the global interconnection of supply-and-demand chains, the food commodity sector is one of the most vulnerable economic sectors to the effects of extreme weather events. The timely identification of abnormal weather and weather risks is a key point in guaranteeing stable supplies by pointing out geographic areas under risk. This risk assessment is directly supporting the accomplishment of United Nation’s sustainable development goal (SDG) 2, particularily by backing the achievement of food security. Realizing the latter has been the main focus during the development of our web application, through supporting stakeholders in the food commodity trading sector in planning advance purchases of supplies. An early planning of purchasing volumes is necessary to prevent disruptions in supply chains and sudden price increases for consumers of final goods.

Over the course of the past years, green spin has been developing, in close exchange with users in the food commodity industry, a web-based application in which data from the Copernicus Climate Change Service (C3S), Copernicus Land Monitoring Service (CLMS), the German Weather Service (“Deutscher Wetterdienst”, DWD), National Oceanic and Atmospheric Administration (NOAA), Global Inventory Monitoring and Modeling System (GIMMS), MODIS and SMAP satellites have been combined to not only provide access to the data but also to extract information to support decision making processes.

Based on the extracted user needs, the systemic knowledge of crop cultivation cycles (mainly wheat, corn and rice) and constant evaluations during the development phase, it has been found that the following parameters contain useful information:

1) Parameters with daily temporal resolution: precipitation (DWD), temperature (DWD), soil water index (CLMS), leaf area index (MODIS), vegetation health index (NOAA), NDVI anomalies (GIMMS), snow cover extent (MODIS) and snow mass (SMAP)
2) Parameters with monthly temporal resolution: temperature 3 months forecasts (C3S), precipitation 3 months forecasts (C3S)

The whole processing pipeline is fully automated and includes downloading, conversion, cleaning of data errors and data extraction. Prepared data are then stored in data bases, checked for integrity and completeness and can be accessed via APIs. So far, data since the year 2000 have been integrated (with exception of soil water index which is only available since 2007). All daily input data are aggregated on administrative levels, ranging from district level (corresponding to “Kreise” in Germany) up to country level, and integrated as interactive maps into the web application. Parameters with a monthly resolution are displayed as continuous vector maps, since they are mostly used as approximation for a quick global assessment of potential medium-range climate developments. This extensive framework has been operational for two years and is continuously evaluated regarding the inclusion of new data.

In addition to data visualization via interactive maps, the data was further used as input for a weather-based risk indicator and the modelling of production, yield and area of specific crops. It was necessary to integrate those additional parameters in order to make the transition from a mere data visualization application to an actively used application in which EO climate and weather data and derived products truly serve as a basis for decision-making for stakeholders in non-EO disciplines.
Other existing solutions like GADAS or geoGLAM provide very good overviews on a number of different parameters and data sets, but are oftentimes difficult to use and interpret due to a high level of complexity. For example, the better part of such portals doesn’t provide analysis tools with which the plethora of displayed parameters and indices can be inter-compared in time as well as in space by the end-user.
Therefore, we developed our web application as a “collaboration framework” with the goal of enabling non-EO users to access EO data and derived information through a solution-oriented approach. On the one hand, our approach does leave out the raw raster data which can be seen as information loss. But on the other hand, the resulting gain in simplicity of interpretation enabling the simplified derivation of information caused by the means of data condensation (spatial aggregation) largely outweighs the perceived loss of information. This approach is based on constant exchange with users, leading to constant adjustments in the application.

One example of a simplified derivation of information is the above-mentioned weather-based risk indicator. It is used to spot “risk areas” on a sub-national scale (first level below country level). The aim of these risk areas is to identify regions in the world where crop areas are under risk due to extreme weather events. Therefore, different existing algorithms have been analyzed to find a representative index which not only provides more information for the risk assessment task but also can be interpreted and understood correctly by non-expert end-users.

The calculation of the risk indicator is based on the computation of the Standard Precipitation Index (SPI) from the National Drought Mitigation Center. Instead of using only precipitation data, additionally temperature and soil water index data were used. Thereby, an index was created which is especially adapted to and targeted on plant growth. Crop production and area statistics are used in order to grade the severeness of the detected risk in an area. The more crop production proportionally present in an area, the higher the severity.
Compared to other existing data portals and applications, the following novel features have been introduced:

• Information is made available for countries and lower administrative units for which data is typically not available (e.g., provincial data for Russia and China)
• Possibility to not only compare time series of different parameters among each other but also to put them directly in relation to crop harvest quantities, enabling the merge of the subjective knowledge of end-users with objective data
• Detection and monitoring of extreme weather events, including their impact on the development of crop production

In conclusion, with this web application we show:
• how EO data can be made accessible for a range of applications (EO- and non-EO related)
• how new parameters from already used or new data providers can be flexibly integrated into the operational data processing pipeline and visualization
• a “best practice” approach how to condense data into useful information with a focus on facilitating comprehension for decision makers from non-EO fields

As an outlook, we are testing the expansion of the presented risk indicator with the integration of temperature and precipitation forecasts as well as population data as an additional measure for assessing severity. These modifications are intended to make the risk indicator more by taking into account the differences between the global (supply chain management) and the local ("direct-to-food" production) context of application.

Atmospheric ozone is an Essential Climate Variable (ECV) monitored in the framework of the Global Climate Observing System (GCOS), among others due to its impact on the radiation budget of the Earth, its chemical influence on other radiatively active species, and its role in atmospheric dynamics and climate. Its importance in the context of climate change has led ECMWF to set up a dedicated procurement of state-of-the-art ozone Climate Data Records (CDRs) to the Climate Data Store (CDS) of the Copernicus Climate Change Service (C3S), mainly in the form of level-3/4 gridded data products. In support, ESA ensures the round-robin selection, reprocessing, and further improvement of the underlying level-2 ozone data products and their validation, and the development of new and multi-spectral ozone CDRs through its Climate Change Initiative project on ECV ozone (Ozone_cci). In order to assess the fitness-for-purpose of the datasets procured to the Copernicus CDS, processes have been established both within the Ozone_cci (L2 data) and C3S (L3/4 data) projects to monitor ozone CDR quality, check compliance with GCOS requirements and WMO rolling review of requirements (RRR), and regularly report key performance indicators. The ozone datasets typically undergo a harmonized and comprehensive quality assessment, including: (a) verification of their information content and geographical, vertical and temporal representativeness against specifications; (b) quantification of their bias, noise and decadal drift, and their dependence on major influence quantities; and (c) assessment of the mutual consistency of CDRs from different sounders.

This work summarizes the past development and the operational status of the data production and quality assessment of the ozone CDRs procured to the CDS. These CDRs consist of ozone column and vertical profile datasets at level-3 (monthly gridded) and level-4 (assimilated), from several nadir and limb/occultation satellite sounders, retrieval systems, and merging schemes (see details on C3S Climate Data Store at https://cds.climate.copernicus.eu/). The quality assessment of these climate-oriented ozone data records is based on multi-decade time series of correlative measurements collected from monitoring networks contributing to WMO’s Global Atmosphere Watch, such as GO3OS, NDACC, and SHADOZ. Correlative measurements are quality controlled, harmonized, and compared to the various satellite CDRs using BIRA-IASB’s Multi-TASTE versatile validation system, following the latest state-of-the-art protocols and tools. Comparison results document the current quality of the CDRs, which may exhibit cyclic errors, drifts, and other long-term patterns reflecting, e.g., instrumental degradation, residual biases between different instruments and changes in sampling of atmospheric variability and patterns. The total ozone column CDRs, covering up to four decades, are found to be stable with respect to the reference measurements at the 0.1 % per decade level. Similarly, most nadir and limb profile CDRs achieve a level of stability that is consistent with what is expected from instrument specifications.

Lakes are a critical natural resource of significant interest to the scientific community, local to national governments, industries and the wider public. Lakes support a global heritage of biodiversity and provide key ecosystem services and they are included in the United Nations’ Sustainable Development Goals committed to water resources and the impacts of climate change. Lakes are also key indicators of local and regional watershed changes, making lakes useful for detecting Earth’s response to climate change. Specifically, lake variables are recognised by the Global Climate Observing System (GCOS) as an Essential Climate Variable (ECV) because they contribute critically to the characterization of Earth’s climate. The scientific value of lake research makes it an essential component of the United Nations Framework Convention on Climate Change (UNFCCC) and the Intergovernmental Panel on Climate Change (IPCC).

The Lakes ECV as defined by GCOS-200 include the following thematic variables:
• Lake water level, fundamental to our understanding of the balance between water inputs and water loss.
• Lake water extent, a proxy for change in glacial regions (lake expansion) and drought in many arid environments. Water extent relates to local climate for the cooling effect that water bodies provide.
• Lake surface water temperature, correlated with regional air temperatures and a proxy for mixing regimes, driving biogeochemical cycling and seasonality.
• Lake ice cover freeze-up in autumn and advancing break-up in spring are proxies for gradually changing climate patterns and seasonality.
• Lake water-leaving reflectance, a direct indicator of biogeochemical processes and habitats in the visible part of the water column (e.g., seasonal phytoplankton biomass fluctuations), and an indicator of the frequency of extreme events (peak terrestrial run-off, changing mixing conditions).
• Lake ice thickness, which provides insight into the thermodynamics of lake ice at northern latitudes in response to changes in air temperatures and on-ice snow mass.

Observing and monitoring precisely and accurately the spatial and temporal variability and trends of the lake thematic variables from local to global scale have become critical to understand the role of lakes in weather and climate, but also for a range of scientific disciplines including hydrology, limnology, biogeochemistry and geodesy. Remote sensing provides an opportunity to extend the spatio-temporal scale of lake observations.

The ESA Lakes_cci dataset presented here includes all the Lake ECV variables, except the lake ice thickness which is in development. The dataset consists of daily observations for each thematic variable over the period 1992-2021. The dataset for each of the thematic variable has been derived from multiple instruments onboard multiple satellites with the compatible algorithms and in an effort to ensure homogeneity and stability over time.

All the thematic variables are reported on a common latitude-longitude grid of about 1km resolution for 2024 lakes distributed globally and covering a wide range of hydrological and biogeochemical regimes. For each of the thematic variables, the observations are accompanied by an uncertainty estimate which makes the dataset particularly suitable for climate applications.

An overview of the thematic variable datasets, their validation, the geographical distribution of the lakes and the way to access the data dataset will be presented together with some major global trends observed in the Lakes ECV.

Lake surface water temperature (LSWT), which describes the temperature of the lake at the surface, is a recognised Essential Climate Variable (ECV) of the Global Climate Observing System (GCOS). It is one of the key parameters determining the ecological conditions within a lake, since it influences physical, chemical and biological processes. LSWT also plays a key component in the hydrological cycle, determining both air-water heat and moisture exchanges. As such, monitoring LSWT globally can be extremely valuable in detecting localised climatic extremes, forewarning authorities to the potential impact of such events on lake ecosystems. But also, operational LSWT observations have potential environmental and meteorological applications for inland water management and numerical weather prediction (NWP) through assimilation.

Through the Copernicus Global Land Surface (CGLOPS) project, we have developed and operationalised a global LSWT dataset that provides a thermal characterization of over 1000 of the world’s largest lakes. The operational LSWT product is generated from brightness temperatures observed by the SLSTR instruments onboard Sentinel3A and Sentinel3B. The dataset is currently based on SLSTR Sentinel3A since June 2016 and both SLSTR Sentinel3A and SLSTR Sentinel3B since August 2020.

LSWT is delivered every 10 days, the period covered starting the 1st, 11th and 21st day of each month and providing a 10-day LSWT average along with uncertainty and quality levels. The LSWTs are mapped to a regular grid of about 1 km resolution. The data are routinely available through the CGLOPS data portal with a latency of three days. As part of the routine monitoring of the product, plots showing comparisons of the most recent LSWT against its climatology for each lake are updated together with the spatial distribution of the LSWTs, allowing for easy detection of anomalous events. Another important aspect of the monitoring is the timeliness and the completeness of the SLSTR data at the time of processing. As such, plots showing the completeness of each 10-day product are made available to show users the amount of data that has been used to generate each product. The LSWTs are regularly validated against in situ measurements covering a large portion of the globe. A simple, interactive web-based platform (http://www.laketemp.net/home_CGLOPS/dataNRT/) has been developed to assist with the exploitation of the near-real time information for each lake covered by the CGLOPS LSWT product and reports detailed information on the validation of the product.

The CCI Open Data Portal has been developed as part of the European Space Agency (ESA) Climate Change Initiative (CCI) programme, to provide a central point of access to the wealth of data produced across the CCI programme. It is an open-access portal for data discovery, which supports faceted search and multiple download routes for all the key CCI datasets. The CCI ODP can be accessed at https://climate.esa.int/data.

The CCI Open Data Portal has been in operation since 2015, and since its inception, has provided access to over 450 datasets and has had more than 50 million file accesses. It consists of two front end access routes for data discovery: a CCI dashboard, which shows the breadth of CCI products available and the time ranges which are covered and can be drilled down to select the appropriate datasets; and a faceted search index, which allows users to search for data over a wider range of characteristics. These are supported at the back end by a range of services provided by the Centre for Environmental Data Analysis (CEDA), which includes the data storage and archival, catalogue and search services, and download servers supporting multiple access routes (FTP, HTTP, OPeNDAP, OGC WMS and WCS). Direct access to the discovery metadata is also available, and can be used by downstream tools to build other interfaces on top of these components e.g., the CCI Toolbox uses the search and OPeNDAP access services to include direct access to data.

In the initial phase of the CCI Open Data Portal, a combination of Earth System Grid Federation (ESGF) search and CEDA’s Catalog Service for Web (CSW), were used to provide the functionality of the portal search. However, using the combination of the two services, and the specialised requirements of ESGF, added complexity, and increased the effort needed to publish data, so the portal was redeveloped in 2019 under the CCI Knowledge Exchange project. In this new phase, the Open Data Portal combines search and data cataloguing using OpenSearch with data serving capacity using Nginx and THREDDS, which has simplified the publication process, and allowed more flexibility when including data. A number of innovations have been made to data serving functionality with the adoption of containers and Kubernetes to provide a scalable data service and the provision of an analysis-ready data cache on JASMIN’s object store using Zarr serialisation of source netCDF files. The latter augments the existing data service to provide access to data for the CCI Toolbox application with data rechunked to provide optimal performance for data analysis queries. Publishing has been further streamlined through two changes. First, the servers providing data download and OPeNDAP services (Nginx & THREDDS) are reading directly from the file system so data appears there as it reaches the CEDA archive. Second, through the use of message passing frameworks (RabbitMQ) and containerised processing scripts, we can generate the metadata needed for search in parallel to the files reaching the archive. In some cases, manual changes are needed to this metadata. These are fed in using configuration files and become part of an automated workflow to re-tag the affected data files, leveraging Continuous Integration pipelines.

A key challenge in the operation of the CCI Open Data Portal comes from the heterogeneity of the different datasets that are produced across the Climate Change Initiative programme, with different scientific areas and different user communities all have differing needs in terms of the format and types of data produced. To this end, the work of the CCI Open Data Portal, also includes maintaining CCI data standards. These standards aim to provide a common format for the data, but necessarily, still leaves considerable breadth in the types of data produced. This provides challenges in providing harmonised search and access services, and solutions have been developed to ensure that every dataset can still be fully integrated into our faceted search services.

In this presentation we will describe the CCI Open Data Portal, recent developments, and the lessons that we have learnt from over six years of operations.

1. Introduction
The ESA-CCI High Resolution (HR) Land Cover(LC) project [1] has focused on the study of the spatial resolution in analyzing the role of the land cover in climate modeling. The project has designed a methodology and developed a processing chain for the production of high resolution land-cover and land-cover change products (10/30 m spatial resolution) by using both optical multispectral images and SAR data. The HRLC Essential Climate Variable (ECV) is derived over long time series of data in the period 1990-2019 by considering sub-continental and regional areas. Images acquired by ESA Sentinel-2, Landsat 5/7/8 multispectral sensors, and Sentinel-1, Envisat and ERS 1/2 SAR sensors have been processed for the generation of the final products. Given the spatial resolution and the long time period, this resulted in a big data problem characterized by a huge amount of images and a very large volume of data that have been considered and processed.
This contribution presents the primary products generated by the project that consist of: (i) HR land-cover maps at subcontinental scale derived in a given target year, (ii) a long-term record of regional HR land cover maps, and (iii) land-cover change maps.

2. Generated Products
The HR land-cover maps at subcontinental level have been generated using time series of images acquired by Sentinel 1 and 2 in 2019 at a resolution of 10m. The processing has been organized to exploit monthly composite of images that can properly represent the seasonality of the classes. With respect to the previous ESA-CCI Land Cover (LC) project [2], the resolution is improved of more than one order of magnitude (from 300m to 10m). Accordingly, the legend of classes has been re-designed to catch the capability of the most recent sensors in capturing smaller objects (e.g., single trees) and their evolution over time. The legend has been defined over 2 levels, where the second one captures the class seasonality, for a total of 20 classes (see figure 1). The HR land-cover maps at subcontinental level behaves as reference static input to the climate models representing the context at high resolution and high quality given the large quantity of available data.

The long-term record of regional HR land cover maps includes 5 maps generated every 5 years in the period 1990-2015. The spatial resolution is of 30m in the regions of interest for the historical analysis (included in but smaller than the regions covered by the sub-continental one). In this time span, the number of yearly-based images available in archives dramatically reduces. This makes the classification problem more challenging. The processing can rely on few images per-year only (in some areas there is only one image or no images) that are thus organized in seasonal or yearly composites depending on data availability. Accordingly, depending on data availability, a higher level legend consistent with the one of the static map has been considered that does not include the seasonal class information when no seasonal information is available (see figure 1).

Land-cover change information is computed yearly at 30 m spatial resolution and is consistent with historical HR land-cover maps. Change information is provided as presence and absence of change, and for changed samples the year of change is provided together with change probability. The change legend considers the climatic most relevant transitions among the possible ones given the LC legend.

All the products are associated with a measure about their uncertainty. The land-cover products also provide information of the second most probable class identified by the classifier on each pixel. This allows to better capture the complexity of the land-cover RCV in input to climate models.

3. Study Areas
The above-mentioned products have been generated over 3 test areas identified by the Climate User Group as of particular interest to study climate change and the related effects in terms of land-cover and land-cover changes. The areas are in three continents involving climate (tropical, semi-arid, boreal) and complex surface atmosphere interactions that have significant impact not only on the regional climate but also on large-scale climate structures. The three regions are in Amazon basin, the Sahel band in Africa and in the northern high latitudes of Siberia as detailed below (see figure 2).

Amazon. This region has been selected due to large deforestation rates, fire drought and agricultural expansion. Those phenomena are potentially associated to large-scale climate impacts and agents of disturbance including losses of carbon storage and changes in regional precipitation patterns and river discharge with some signs of a transition to a disturbance-dominated regime. An example of LC maps for Amazon is given in Figure 3.

Africa. This region is associated to Sahel band including West and East Africa, which is a complex climatic region which experiences severe climatic events (droughts and floods) for which the future predictions are very uncertain. In this area HRLC impact can be evaluated on better modeling the position and seasonal dynamics of the monsoons (the West African and the Indian ones) and surface processes; and on the explanation of the role of El Nino in the initiation of dramatic drought events (eastern part of the Sahelian band).

Siberia. The third region is expected to be strongly affected by climate changes (polar amplification). Mapping LC changes can document the displacement of the forest-shrubs-grasslands-transition zone to the north and the impact on the carbon stored in permafrost, which in turn will affect long-term terrestrial carbon balance and ultimately climate change.

The generated products have been systematically validated both qualitatively and quantitatively (in terms of overall, producer and user accuracy), and intercomparison analysis has conducted with other land-cover products. Sample collection for quantitative analysis has conducted by photointerpretation on very high resolution images (higher than the 10/30m resolution of products) and intercomparison relies on other existing maps for the considered study areas. The products and the related validation will be presented at the symposium.

References
[1] L. Bruzzone et al, "CCI Essential Climate Variables: High Resolution Land Cover,” ESA Living Planet Symposium, Milan, Italy, 2019.
[2] P. Defourny et al (2017). Land Cover CCI Product User Guide Version 2.0. [online] Available at: http://maps.elie.ucl.ac.be/CCI/viewer/download/ESACCI-LC-Ph2-PUGv2_2.0.pdf

List of the other HRLC team members: M. Zanetti (FBK), C. Domingo (CREAF); K. Meshkini (FBK), C. Lamarche (UCLouvain), L. Agrimano (Planetek), G. Bratic (PoliMI), P. Peylin (LSCE), R. San Martin (LSCE), V. Bastrikov (LSCE), P. Pistillo (EGeos), I. Podsiadlo (UniTN), G. Perantoni (UniTN), F. Ronci (eGeos), D. Kolitzus (GeoVille), T. Castin (UCLouvain), L. Maggiolo (UniGE), D. Solarna (UniGE).

During the last decades, several sensors were launched that allowed the study of wildfires from space at a global scale. They provide information on active fires, area burned, and the regeneration of the vegetation after the fire event. One of the key variables to assess the impact of wildland fires on climate, in terms of greenhouse gasses and particulate matter emissions, is to know the area of the vegetation burned during the fires.
To address this need, the ESA CCI Fire Disturbance project (FireCCI) has developed in the last years a suite of burned area (BA) products based on different sensors, creating a database spanning from 1982 to 2020. These products, apart from providing information on burned area, also include ancillary information related to the uncertainty of the detection, the land cover affected (extracted from the Land Cover CCI product), and the observational limitations of the input data. All products supply information in monthly files, and are delivered at two spatial resolutions: pixel (at the original resolution of the surface reflectance input data) and grid (at a coarser resolution and specifically tailored for climate researchers).
The dataset with the longest time series is the FireCCILT11 product, based on AVHRR information obtained from the Land Long-Term Data Record (LTDR) version 5, and spanning from 1982 to 2018 at a global scale (Otón et al. 2021). The pixel product has a spatial resolution of 0.05 degrees (approx. 5 km at the Equator), and provides information on the date of the fire detection, the confidence level of that detection, the burned area in each pixel, and an ancillary layer with the number of observations available for the detection. The grid product, at a resolution of 0.25 degrees, summarizes the data of the pixel product for each grid cell, and includes layers corresponding to the sum of burned area, the standard error, and the fraction of burnable area and observed area in each cell. FireCCILT11 is the global BA product with the longest time-series to date.
Another global product, but with a higher spatial resolution, is the FireCCI51, whose algorithm uses MODIS NIR surface reflectance at 250 m spatial resolution and active fires as input (Lizundia-Loiola et al. 2020). This product has a time series of 20 years (2001 to 2020), and it is the global burned area product with the highest resolution currently available. The pixel product includes layers corresponding to the date of detection, the confidence level and the land cover burned, while the grid product, at 0.25-degree resolution, contains the same information as FireCCILT11, and also includes layers of the amount of burned area for each land cover class.
As part of our effort to extend this burned area information into the future, the FireCCI project has recently developed a new algorithm to detect BA using the SWIR bands of the Sentinel-3 SLSTR sensor, extracted from the Synergy (SYN) products developed by ESA. This product, called FireCCIS310, takes advantage of the improved BA detection capacity of the SWIR bands, which has allowed to detect approx. 20% more burned area than the previous global datasets, and with an increased accuracy. FireCCIS310 is currently available for the year 2019, but will be extended into the future. It supplies the same layers as FireCCI51, but at a spatial resolution of 300 m for the pixel product.
Finally, a specific dataset has been created for sub-Saharan Africa, where more than 70% of the total global burned area occurs. This product, called Small Fire Dataset (SFD) uses surface reflectance from the Sentinel-2 MSI sensor at 20 m spatial resolution, complemented with active fire information (Roteta et al. 2019). Version 1.1 of this dataset (FireCCISFD11) covers the year 2016 and is based on Sentinel-2 A data. It includes the same pixel and grid layers as the FireCCI51 product. The newer version 2.0 (FireCCISFD20) has been processed for the year 2019, and takes advantage of the additional data provided by Sentinel-2 B, duplicating the input data amount and temporal resolution. The grid version of this product has a spatial resolution of 0.05 degrees, as suggested by the climate researchers. Due to its higher spatial resolution, this product detects 58% more BA than FireCCI51 for 2016, and 82% in 2019. The vast majority of this additional BA is due to the improved detection of small burned patches, not detectable with moderate resolution sensors.
The increase of burned area detection has a direct impact on climate research, as more vegetation burned means more atmospheric emissions. Carbon emissions from FireCCISFD11, for instance, are between 31 and 101% higher than previous estimates for Africa, and represent about 14% of global CO2 emissions from fossil fuels (Ramo et al. 2021). The BA algorithms and products developed by FireCCI are, therefore, contributing to this line of research, providing new and more accurate information to the climate community.

References:
Lizundia-Loiola, J., Otón, G., Ramo, R., Chuvieco, E. (2020) A spatio-temporal active-fire clustering approach for global burned area mapping at 250 m from MODIS data. Remote Sensing of Environment 236, 111493, https://doi.org/10.1016/j.rse.2019.111493
Otón, G., Lizundia-Loiola, J., Pettinari, M.L., Chuvieco, E. (2021) Development of a consistent global long-term burned area product (1982–2018) based on AVHRR-LTDR data. International Journal of Applied Earth Observation and Geoinformation 103, 102473. https://doi.org/10.1016/j.jag.2021.102473
Ramo, R., Roteta, E., Bistinas, I., Wees, D., Bastarrika, A., Chuvieco, E. & van de Werf, G. (2021) African burned area and fire carbon emissions are strongly impacted by small fires undetected by coarse resolution satellite data. PNAS 118 (9) e2011160118, https://doi.org/10.1073/pnas.2011160118
Roteta, E., Bastarrika, A., Padilla, M., Storm, T., Chuvieco, E. (2019) Development of a Sentinel-2 burned area algorithm: Generation of a small fire database for sub-Saharan Africa. Remote Sensing of Environment 222, 1-17, https://doi.org/10.1016/j.rse.2018.12.011

The dynamics of the Sea Ice affects the global Earth system. Changes in polar climate have an impact across the world, affecting lives and livelihoods, and regulating the climate and the weather. CRiceS (Climate relevant interactions and feedback: the key role of sea ice and snow in the polar and global climate system) is a recent European project aiming at understanding the role of ocean-ice/snow-atmosphere interactions in polar and global climate. The main objective of CRiceS is to deliver improved understanding of the physical, chemical, and biogeochemical interactions within the Ocean/Ice/Atmosphere system, new knowledge of polar and global climate, and enhanced ability of society to respond to climate change.

One of the variables that plays a key role for a better understanding of the ocean/Ice/atmosphere dynamics is Sea Surface Salinity (SSS). SSS allows monitoring changes in sea ice by means of the study of their positive anomalies (associated to sea ice formation and evaporation) and negative anomalies (associated with melting and precipitation). The acquisition of in situ salinity measurements in polar regions is very complicated because of the distance and the extreme weather conditions. Therefore, measurements acquired by satellites become the unique way of having a continuous and synoptic monitoring of the sea surface salinity in polar regions.

Acquisitions of L-band satellite SSS in polar regions, and particularly those by ESA SMOS mission, are hampered by the decrease of sensitivity of brightness temperatures to SSS in cold waters. Recently, these difficulties have been overcome in a dedicated project from ESA (Arctic+ Salinity) over the Arctic Ocean, leading to satellite SSS measurements with enough quality to address many scientific studies.

However, in the Southern Ocean, since the salinity variability is not as large as in the Arctic Ocean, current quality of L-band brightness temperatures does not always allow assessing the seasonal and interannual salinity dynamics of the region. For this reason, reducing brightness temperature errors in this region is one of the major requirements to obtain SSS of enough quality to address scientific studies here.

In the framework of the ESA regional initiative called SO-FRESH, new and enhanced algorithms to reduce the brightness temperature errors have been applied for generating a new SMOS SSS regional product in the Southern Ocean. In this work, we will use the enhanced SMOS SSS product generated in this project and we will present a preliminary quality assessment by: i) comparing with in situ measurements; ii) analysing the uncertainty estimations by means of correlated triple collocation analysis; iii) analysing the seasonal behaviour by using harmonic analysis; and iv) assessing its effective spatial resolution with singularity and spectral analysis. Finally, we will show the capability of this product of improving the description of the Ocean Ice and Atmospheric system in numerical models which is one of the main scientific objectives of CRiceS.

The Copernicus Climate Change Service (C3S) is one of the six thematic information services provided by the Copernicus Earth Observation Programme of the European Union (EU). C3S, which is implemented by ECMWF on behalf of the European Commission, provides past, present and future Climate Data Record (CDR) and information on a range of themes, freely accessible through the Climate Data Store (CDS). It benefits from a sustained network of in-situ and satellite-based observations, re-analysis of the Earth climate and modelling scenarios, based on a variety of climate projections.

Within the Land Biosphere component of C3S, satellite-based observations are used to provide the longest possible, consistent and mature products at the global scale for the following Essential Climate Variables (ECVs): Surface albedo, Leaf Area Index (LAI), the fraction of Absorbed Photosynthetic Active Radiation (fAPAR), Land Cover (LC), Fire, Burnt Areas (BA), and Fire Radiative Power (FRP). State-of-the-art algorithms that respond to GCOS requirements are used and the product quality assurance follows the protocols, guidelines and metrics defined to be consistent with the Land Product Validation (LPV) group of the Committee on Earth Observation Satellite (CEOS) for the validation of satellite-derived land products.

To reach this goal, the following approach is proposed.(i) consolidate the CDR and secure continuation of the products by moving towards Copernicus mission Sentinel-3 as primary data source,(ii) make an important steps towards cross-CDR consistency by harmonizing the pre-processing for all CDRs (atmospheric correction and pixel classification) and (iii) apply an extensive quality with other existing datasets to ensure high quality of the delivered data.

The Belgian VITO Remote Sensing institute is leading the consortium of eight European partners providing the C3S service from 2016. In the new phase of the C3S service all ECVS will use Sentinel 3 and the adaptations will be done in consistency with previous products. The surface albedo V3 data set will be extended in time based on Sentinel 3 OLCI/SLSTR dataset. Improving Land Cover and Burnt Area will be achieved by (i) extending the already existing BA and Land Cover CDRs and ICDRs from the respective projects, (ii) adapting them to Sentinel-3 SLSTR and OLCI sensors, (iii) benefitting from the harmonised pre-processing tools (pixel identification and AC LUTs), and (iv) incorporating these into the processing chains for fully operational and agile production lines.

The Burned Areas product created ad-hoc for C3S, based on the algorithm developed by ESA Fire_CCI, but adapted to Sentinel-3A and B will continue to be processed. A major advancement will be achieved by switching from MODIS based active fire maps to active fire from Sentinel 3, once this is available from the Copernicus ground segment, expected in early 2022. The Active Fire and Fire Radiative Power products will be continued in the service using only Sentinel-3 night-time data as input. ESA has announced that the daytime fire products will be available from the Copernicus ground segment from late 2021/early 2022. Once these data are available, an update is planned for C3S Level2/3 products. Availability of daytime active fires and fire radiative power is highly needed, and, for example, our Fire BA products will be enabled to switch from ageing MODIS to in-house Sentinel 3 auxiliary data thanks to this.

The high quality and maturity of the generated ECV datasets make them a reliable indicator of long-term climate predictions and will contribute information to the annual European State of the Climate report. More detailed information about the ongoing activities and results of the Lot 5 C3S project will be shared at the Living Planet Symposium.

The Orbiting Carbon Observatory 3 (OCO-3) was installed on the International Space Station (ISS) Japanese Experiment Module – External Facility (JEM-EF) in May 2019. From that vantage point, it is using the flight spare instrument from OCO-2 to collect observations of reflected sunlight that are analyzed to return additional estimates of the CO2 dry air mole fraction, XCO2, and solar-induced chlorophyll fluorescence (SIF). The ISS JEM-EF is a highly sought-after resource, so missions installed there are planned for a limited lifetime of typically three years. OCO-3 began routine operations in August 2019 and has an operating extension beyond the nominal three years to at least January 2023. Here, we will present the mission status, including instrument performance, key mission events, data collection statistics and highlights of the science findings of the mission to date. To prepare for the end of the mission, the team will develop a final data product, Version 11, and complete the mission documentation. Details of the end of mission plans, how they fit with the OCO-2 mission and how the data collected is advancing monitoring of urban/local emissions will be discussed.

Concentrations of atmospheric methane (CH4), the second most important greenhouse gas, continue to grow. In recent years this growth rate has increased further (2020: +14.7 ppb), the cause of which remains largely unknown. Accurate estimates of CH4 emissions are key to better understand these observed trends and help implement efficient climate change mitigation policies. New methane observations from the TROPOMI instrument provide unprecedented spatiotemporal constraints on these emissions. Here, we present preliminary results from a new inversion system based on the ECMWF Integrated Forecasting System (IFS) ,which assimilates observations within a 24-hour window cycled 4D-variational algorithm. Specificities of this system include the use of a high-resolution transport model (~9km) combined with online data assimilation (i.e., joint optimization of meteorological and atmospheric composition variables) that provides consistent treatment of atmospheric transport errors. The performance of the system is illustrated by comparing posterior atmospheric concentrations with independent observations, as well as by evaluating posterior emission estimates for regional and point source case studies previously analyzed in the literature. The largest national disagreement found between prior (63.1 Tg yr-1) and posterior (59.8 Tg yr-1) CH4 emissions is from China, mainly attributed to the energy sector. Emissions estimated form our global system agree well with previous basin-wide regional studies and point source specific studies. Emission events (leaks/blowouts) >10 t hr-1 were detected, but without accurate prior uncertainty information, were not well quantified. Our results suggest that global anthropogenic CH4 emissions for 2020 were 5.7 Tg yr-1 (+1.6%) higher than for 2019, mainly attributed to the energy and agricultural sectors. Regionally, the largest increases were seen from China (+2.6 Tg yr-1, 4.3%), with smaller increases from India (+0.8 Tg yr-1, 2.2%) and Indonesia (+0.3 Tg yr-1, 2.6%). Plans to further develop the global IFS inversion system and to extend the 4D-Var window-length using a hybrid ensemble-variational method will also be presented.

Methane (CH4) is the second most important greenhouse gas, of which more than 60 % CH4 is released through human activities. Satellite observations of CH4 provide an efficient way to analyze its variations and emissions. The TROPOspheric Monitoring Instrument (TROPOMI) onboard the Sentinel 5 Precursor (S5-P) satellite measures CH4 at a high horizontal resolution of 7 × 7 km2, showing the capability of identifying and quantifying the sources at a local to regional scale. The Middle East is one of the strong CH4 hotspot regions in the world. However, it is difficult to estimate the emissions here because several sources are located near the coast or in places with complex topography, where the satellite observations are often of reduced quality. We use the WMF-DOAS XCH4 v1.5 product, which has good spatial coverage over the ocean and mountains, to better estimate the emissions in the Middle East.
The divergence method of Liu et al., (2021) has been proven to be a fast and efficient way to estimate CH4 emissions from satellite observations. We have improved our method by comparing the fluxes in different directions for better background corrections over areas with complicated topographies. The performance of the updated algorithm was tested by comparing the estimated emissions from a 1-month WRF-CMAQ model simulation with its known emission inventory over the Middle East. The CH4 emissions based on TROPOMI XCH4 are then derived on a 0.25° grid for 2019 and 2020. With the WMF-DOAS product, sources from oil/gas platforms over the Persian Gulf and sources on the west coast of Turkmenistan become clearly visible in the emission maps. Sources in the mountain areas of Iran are also identified by our updated divergence method. The locations of fossil fuel related NOX emissions usual overlap with CH4 emissions as can be seen in the CAMS bottom-up inventory. Therefore, we have compared our CH4 emission inventory with the emissions derived from TROPOMI observed NO2, in order to gain more insight into the source of the emissions, especially concerning the oil/gas industry in the region.

The Copernicus Anthropogenic Carbon Dioxide Monitoring (CO2M) mission is the first operational space-base system aimed at collecting data in support of systems for global monitoring and verification of CO2 emissions. This will require sampling major emission areas (including plumes from point sources and cities) with high coverage and sufficiently high accuracy including regions with enhanced aerosol loadings.

CO2M has been designed to meet these objectives by carrying an imaging spectrometer for CO2 measurements (CO2I) together with a multi-angle polarimeter (MAP) for co-located aerosol information. The underlying assumption is that the MAP instrument can provide a detailed aerosol characterization for the CO2 retrieval and thus allows to reduce critical aerosol-related uncertainties.

Making use of aerosol information from the MAP instrument will require to develop new approaches for the CO2 retrieval. We have developed a sequential approach where first aerosol properties are retrieved from the MAP measurements which are then used as input for the CO2 retrieval from the CO2I observations. This new retrieval brings together the Generalized Retrieval Aerosol and Surface Properties (GRASP) for the MAP retrieval with the University of Leicester (UoL) full-physics retrieval for the CO2 retrieval.

In this presentation, we will give a description of the sequential MAP-CO2 retrieval for CO2M and present a characterisation of the retrieval approach based on global simulations of realistic atmospheric scenarios. The presentation will conclude with an outlook towards further development needs.

Climate information is essential for monitoring the success of our efforts to reduce greenhouse gas emissions that contribute to climate change, as well as for promoting efforts to increase energy efficiency and to transition to a carbon-neutral economy. The WMO Integrated Global Observing System (WIGOS) promotes network integration and partnership outreach, and engages the regional and national actors essential for successful integration of these systems. The WIGOS Vision for 2040 outlines the ground and space-based capabilities that are required in 2040 to deliver the observations required. These data and observations rely on the Global Climate Observing System (GCOS), which maintains the requirements of Essential Climate Variables (ECVs), and support additional observational needs that are required to systematically observe Earth`s changing climate and is such underpin climate research, services and adaptation measure.
The 2021 Extraordinary World Meteorological Congress approved the new WMO Unified Data Policy, along with two other sweeping initiatives – the Global Basic Observing Network (GBON) and the Systematic Observations Financing Facility (SOFF) – to dramatically strengthen the world’s weather and climate services through a systematic increase in much-needed observational data and data products from across the globe. Approval of the Unified Data Policy provides a comprehensive update of the policies guiding the international exchange of weather, climate and related Earth system data between the 193 Member states and territories of WMO. The new policy reaffirms the commitment to the free and unrestricted exchange of data, which has been the bedrock of WMO since it was established more than 70 years ago.

The Global Basic Observing Network (GBON) is a landmark agreement offering a new approach in which the basic surface-based observing network is designed, defined and monitored at the global level. It paves the way for a radical overhaul of the international exchange of observational data, which underpin all weather, climate and water services. This becomes increasingly important also for climate and greenhouse gas monitoring, when the ground-based and space-based components are used in an integrated fashion. Data from programmes like the WMO Global Atmospheric Watch (GAW) and Integrated Global Greenhouse Gas Information System are key for a comprehensive analysis and monitoring of greenhouse gases and climate and will play an increasingly important role supporting satellite observing systems providing ground-truth and much required data for satellite calibration and validation activities. The new WMO Data policy, and GBON, provide the tools and mechanisms to further evolve these systems to meet future needs for a comprehensive climate, greenhouse gas and carbon monitoring system.

This presentation will give an overview of the above elements and how WMO and GCOS support greenhouse gas and climate monitoring activities and facilitates and leverages access to ground-based observations in response to global needs.

In early 1990s, a European consortium led by French and Greek universities and geophysical observatories initiated an institution of long-term observation in the western Gulf of Corinth, Greece, named the Corinth Rift Laboratory (CR, http://crlab.eu). Its principal aim is to better understand the physics of the earthquakes, their impact and the connection to other related phenomena such as tsunamis or landslides.
The Corinth Rift, is one of the narrowest and fastest extending continental regions worldwide. Its western termination was selected as the study area with the criterion of its high seismicity and strain rate. The cities of Patras and Aigio, as well as other towns were destroyed several times since the antiquity by earthquakes and, in some cases, by earthquake-induced tsunamis. The historical earthquake catalogue of the area reports five to ten events of magnitude larger than 6 per century. Episodic seismic sequencies are often. Over the past two decades, a dense array of permanent sensors was established in the CRL, gathering 80+ instruments, the majority of them being acquired in real time.
The CRL is nowadays one of the Near Fault Observatory (NFO) of the European Plate Observing System (EPOS, https://www.epos-eu.org/tcs/near-fault-observatories) and the only one with international governance.
With the development of synthetic aperture radar interferometry (InSAR) and high-resolution optical imagery space missions, remote sensing occupies an increasingly important place in the observatory. Space observations, especially those from InSAR, contain unique, dense and global information that cannot be obtained through field observations. Although low Earth orbit satellites cannot provide continuous real-time observations, the time lag can be sufficiently short for the space products to be useful for monitoring needs.
For the observation of the CRL observatory, the European Space Agency’s Geohazards Exploitation Platform (GEP) gathers, in a well-organized manner, products routinely made by different services, with a double benefit for the observatory: (1) computational resources and algorithms hosted and maintained by the service provider and (2) capability to elaborate solutions with different services for greater confidence and robustness.
An additional advantage is the didactic and user friendly design of the GEP and secondary education teachers. This experiential summer school is tailored to teach in this natural laboratory and in the field the major components and theoretical background of the observations performed in the NFO. Space observations occupy an important role in the school, with the presence of experts from space agencies and the GEP consortium. The participants have the opportunity to analyze the space data directly in the field, in front of the in-situ instruments as well as in front of geological and other objects of interest. The CRL-School is particularly relevant to the activities of ESA’s European Space Education Resource Office (ESERO) network of currently twenty offices in the ESA member states, focusing on strengthening Science, Technology, Engineering, and Mathematics (STEM) and Space Education in primary and secondary education.

Carbon emissions related to fossil fuel tend to come from localised sources, with urban areas in particular contributing more than 70% of global emissions. In the future, the proportion of the world's population living in cities is expected to continue to rise, resulting in a shift towards an even greater contribution towards fossil fuel related emissions originating from urban areas. Cities are also the focal point of many political decisions on mitigation and stabilisation of carbon emissions, often setting more ambitious targets than national governments (e.g. through the C40 group of cities around the world). For example, the Mayor of London has set the ambitious target for London to be a zero carbon city by 2050. If we want to devise robust, well informed climate change mitigation policies, we need a much better understanding of the carbon budget for cities and the nature of the diverse emission sources within them, underpinned by new approaches that allow verification and optimisation of city carbon emissions and their trends. New satellite observations of CO2 from missions such as OCO-3, MicroCarb and CO2M, especially when used in conjunction with ground-based sensor networks, provide a powerful and novel capability for evaluating and eventually improving existing CO2 emission inventories.

In April 2021 we set up a ground-based measurement network comprising three sites, located upwind, downwind and in the centre of London, using portable greenhouse gas (CO2, CH4, CO) column sensors (Bruker EM27/SUN spectrometers) together with UV/VIS MAX-DOAS spectrometers (NO2). The instruments have so far operated continuously over the course of one year, which we have achieved by automating the sensors and housing them inside weatherproof enclosures. The data we have acquired from the network will not only allow us to critically assess the quality of satellite observations over urban environments, but also to derive data-driven emission estimates using a measurement-modelling framework. Here we will show and discuss findings from our first year of greenhouse gas column observations over London.

Limiting global warming to below 2 degrees Celsius as agreed upon in the Paris agreement requires substantial reductions in fossil fuel emissions. The transparency framework for anthropogenic carbon dioxide (CO2) emissions of the Paris Agreement is based on inventory-based national greenhouse gas emission reports, which are complemented by independent estimates derived from atmospheric CO2 measurements combined with inverse modelling. Such a Monitoring and Verification Support (MVS) capacity is planned to be implemented as part of the EU’s Copernicus programme, however, its ability to constrain fossil fuel emissions to a sufficient extent has not yet been assessed. The CO2 Monitoring (CO2M) mission, planned as a constellation of satellites measuring column-integrated atmospheric CO2 concentration (XCO2), is expected to become a key component of an MVS capacity.

Here we provide an assessment of the potential of a Carbon Cycle Fossil Fuel Data Assimilation System using synthetic XCO2 and other observations to constrain national fossil fuel CO2 emissions for an exemplary 1-week period in 2008 at global scale. We find that the system can provide useful weekly estimates of country-scale fossil fuel emissions independent of national inventories. When extrapolated from the weekly to the annual scale, uncertainties in emissions are comparable to uncertainties in inventories, so that estimates from inventories and from the MVS capacity can be used for mutual verification.

We further demonstrate an alternative, synergistic mode of operation, which delivers a best emission estimate through assimilation of the inventory information as an additional data stream. We show the sensitivity of the results to the setup of the CCFFDAS and to various aspects of the data streams that are assimilated, including assessments of surface networks, the number of CO2M satellites flying in constellation, and the assumed uncertainties in the XCO2 measurements. We also assess the impact of additional observational data streams such as radiocarbon in CCFFDAS on constraining fossil fuel emissions.

Anthropogenic emissions of well-mixed greenhouse gases are currently the main drivers of tropospheric warming. Among the well-mixed greenhouse gases methane (CH4) and carbon dioxide (CO2) are the most important contributors. To limit the global warming, emissions of CH4 and CO2 must be reduced, and reduction claims need to be monitored. Additionally, knowledge of especially CH4 emission sources like landfills, oil, gas and coal production has to be expanded. During the last few years several different satellite sensors demonstrated anthropogenic greenhouse gas emissions detection and/or quantification at various spatial scales and spatial resolution, but there is a lack of airborne systems for emission characterization as well as for validation and verification of the new satellite data. In this context, University of Bremen started the development of a new generation of airborne imaging spectrometer systems for accurate mapping of atmospheric greenhouse gas concentrations (CO2, CH4) based on more than ten years of experience with operating the MAMAP airborne system. The first sensor - in a series of three – is the MAMAP2D-Light (M2DL) instrument. M2DL is a relatively light weight (~42 kg) single channel imaging spectrometer covering absorption bands of CO2 and CH4 between ~1575 and ~1700 nm with a spectral resolution of ~1.1 nm. The instrument is designed to fit into the under-wing pod of a motor glider aircraft (Diamond HK36 TTC-ECO) of the Jade University of Applied Sciences in Wilhelmshaven. At a typical flight altitude of ~1500 m the instrument samples 28 ground scenes across the ~600 m wide swath with a single ground sampling size of approximately 20 m across x 3 m along the flight track. Successful test flights were performed in 2021. While designed to detect and quantify CO2 and CH4 emissions from point sources, it additionally serves as precursor and demonstrator for the larger 2-channel imaging spectrometer MAMAP2D, which is currently being built, as well as the planned ESA CO2M airborne demonstrator.
MAMAP2D (M2D) – currently under construction – is a two channel imaging spectrometer covering the O2A band and the absorption bands of CO2 and CH4 between ~1590 and ~1690 nm with a spectral resolution of < 0.4 nm. The instrument is designed to fit into the cabin of different types of aircraft (pressurised and non-pressurised). At a flight altitude of ~1500 m the instrument samples 37 ground scenes across the ~ 670 m wide swath with a single ground sampling size of approximately 18 m across x 7 m along the flight track.
The third sensor in this series will be the CAMAP2D (Carbon And Methane mAPper 2D), which emerged from ESA’s CO2M airborne demonstrator activities. CAMAP2D will be built for ESA by adding a 2 µm channel to MAMAP2D and further modifications of MAMAP2D to reach the ambitious performance goals for CO2 monitoring.
In this presentation we summarise the status and perspective of the new generation of airborne GHG imaging systems. This will include performance estimates, data analysis strategies as well as initial results from M2DL measurement flight targeting the CO2 emission plume of the coal-fired power plant Jänschwalde in Germany in June 2021. Future applications for emission characterization, satellite data validation and airborne data driven science studies in support of satellite data products from S5P, S5 and CO2M as well as from hyperspectral (PRISMA, ENMAP, CHIME) and spatial very high resolution imagery (WV3, Sentinel 2) will be discussed.

The Imaging and Rapid-scanning mass spectrometer (IRM) onboard Swarm-E frequently measures enhanced minor ionospheric species (N+, NO+, N2+, O2+) at auroral latitudes during both storm and quiet times. With their occurrence frequency peaking in the pre-midnight sector, these ions are thought to be the product of both auroral electron impact ionization and thermospheric expansion. These ions have been measured in ion upflows and downflows and could therefore impact the overall vertical transport and coupling processes in the auroral ionosphere. The dissociative recombination of the measured molecular ions likely constitutes a non-negligible source of hot oxygen atoms, affecting the thermospheric mass density and temperature. Furthermore, the different energy dependence of charge exchange with H between these species could impact the dynamics of storm and substorm recovery. We present new Swarm-E ionospheric composition and velocity measurements and discuss their possible implications in the context of upcoming missions.

The Low Frequency Array (LOFAR) is designed to observe the early universe at radio wavelengths. When radio waves from a distance astronomical source traverse the ionosphere, structures in this plasma affect this signal. The high temporal resolution available (~100 ms), the large range of frequencies observed (10-80 MHz & 120-240 MHz) and the large number of receiving stations (currently 52 across Europe) mean that LOFAR can observe the effects of the midlatitude ionosphere in a level of detail never seen before.

On the 14th July 2018 LOFAR stations across the Netherlands observed Cygnus A between 17:00 UT and 18:00 UT. At approximately 17:40 UT a deep fade in the intensity of the received signal was observed, lasting some 15 minutes. Immediately before and after this deep fade rapid variations of signal strength were observed, lasting less than five minutes. This structure was observed by multiple receiving stations across the Netherlands. It evolved in time and in space. It also exhibited frequency dependent behaviour.

The geomagnetic conditions at the time of the observation were quiet, as were the solar conditions. It is suggested that this structure is driven by a source within the Earth system. Observations from lower in the atmosphere are used to identify possible drivers.

The NanoMagSat mission, currently under development in the context of the ESA Scout missions, will consist of a constellation of three nanosatellites combining two 60° inclined and one polar orbits, all at initial altitude of about 570 km. The mission will target investigations of the Earth’s magnetic field and ionospheric environment. Each satellite will carry identical payloads and include a miniaturized High Frequency Magnetometer (HFM) providing three component measurements of the magnetic field at a cadence of 2,000 samples per second. Here, we investigate the possibility of taking advantage of these future measurements, also using modern analysis techniques, to investigate polarization and propagation properties of two important classes of electromagnetic waves.

Equatorial noise is a natural electromagnetic emission generated by instability of ion distributions in the magnetosphere. These waves, which can also interact with energetic electrons in the Van Allen radiation belts, have been shown to propagate radially downward to the low-Earth orbit, thanks to previous measurements from the DEMETER spacecraft. Such waves have been observed at frequencies both below and above the local proton cyclotron frequency as a superposition of spectral lines from different distant sources. Changes in the local ion composition encountered by the waves during their inward propagation cause well identifiable cutoffs in the wave spectra, which can provide valuable information on the ionospheric plasma.

A second class of electromagnetic waves, also worthy of investigations, are nonlinear whistler mode chorus and chorus-like emissions known for their ability to locally accelerate electron in the outer radiation belt to relativistic energies and to cause losses of electrons from the radiation belts by their precipitation in the atmosphere. A divergent propagation pattern of waves at chorus frequencies has previously been reported at subauroral latitudes. The waves propagated with downward directed wave vectors, which were slightly equatorward inclined at lower magnetic latitudes and slightly poleward inclined at higher latitudes. Reverse ray tracing indicated a possible source region near the geomagnetic equator at a radial distance between 5 and 7 Earth radii. Detailed measurements of the Cluster spacecraft have already shown chorus propagating outward from this source region. The time-frequency structure and frequencies of chorus observed by Cluster along the reverse ray paths suggests that low altitude observations could indeed possibly be made by NanoMagSat, which would correspond to a manifestation of natural magnetospheric emissions of whistler mode chorus.

Introduction
HYDROCOASTAL is a two year project funded by ESA, with the objective to maximise exploitation of SAR and SARin altimeter measurements in the coastal zone and inland waters, by evaluating and implementing new approaches to process SAR and SARin data from CryoSat-2, and SAR altimeter data from Sentinel-3A and Sentinel-3B. Optical data from Sentinel-2 MSI and Sentinel-3 OLCI instruments will also be used in generating River Discharge products.
New SAR and SARin processing algorithms for the coastal zone and inland waters will be developed and implemented and evaluated through an initial Test Data Set for selected regions. From the results of this evaluation a processing scheme will be implemented to generate global coastal zone and river discharge data sets.
A series of case studies will assess these products in terms of their scientific impacts.
All the produced data sets will be available on request to external researchers, and full descriptions of the processing algorithms will be provided

Objectives
The scientific objectives of HYDROCOASTAL are to enhance our understanding of interactions between the inland water and coastal zone, between the coastal zone and the open ocean, and the small scale processes that govern these interactions. Also the project aims to improve our capability to characterize the variation at different time scales of inland water storage, exchanges with the ocean and the impact on regional sea-level changes

The technical objectives are to develop and evaluate new SAR and SARin altimetry processing techniques in support of the scientific objectives, including stack processing, and filtering, and retracking. Also an improved Wet Troposphere Correction will be developed and evaluated.

Presentation
The presentation will describe the different SAR altimeter processing algorithms that are being evaluated in the first phase of the project, and present results from the evaluation of the initial test data set. It will focus particularly on the performance of the new algorithms over inland water.

Soil Moisture derivation from Satellite Radar Altimetry has been pursued over the past ten years with a view to augmenting the observability in terms of space-time sampling, resolution and dynamic range. The basis of this technique involves crafting DRy EArth Models (DREAMs), which model the response of a completely dry surface to nadir illumination at Ku band. Initially developed over desert and semi-arid terrain, where DREAM hydrological content was primarily restricted to salars and dry river courses, DREAM crafting is now being extended to wetter areas.
This paper addresses the following questions:
1) Under what conditions can radar altimeters measure surface soil moisture? Can DREAMs be crafted over river basins?
2) What hydrology information is encoded in river DREAMs?
3) What can Sentinel-3 tell us about deployment of the new generation of satellite radar altimeters in recovery of soil moisture signals?
4) With the spatial and temporal sampling constraints of current and past altimeters, where are these data valuable?

Data from Sentinel-3A, CryoSat-2, EnviSat, Jason 1/2 and ERS1/2, together with a database of over 86000 graded River and Lake time series, are analysed to investigate the feasibility of DREAM crafting over river basins.
In this paper, results are presented over 15 regions where DREAMs have been constructed. DREAMs are crafted from multi-mission satellite altimeter data and imaging data, informed by ground truth. Current DREAMs have a spatial resolution of 10 arc seconds and a typical dynamic range of order 50dB. They are configured such that a 10dB increase in one pixel corresponds to the change from desiccated to fully saturated surface. Scaling altimeter backscatter for each mission to the DREAMs allows direct estimation of surface soil moisture.
In desert DREAM areas, small seasonal soil moisture signals were successfully retrieved. The first DREAM with significant hydrological content was developed over the Kalahari desert. Altimeter derived soil moisture estimates were generated and compared with external validation data, including the ESA CCI dataset (Dorigo et al., 2017). Good agreement was obtained.
To progress this approach, it was decided to trial the DREAM methodology to craft first generation models over the Congo and Amazon basins.

For these models, the first requirement was to mask off areas of permanent or seasonal inundation. The first test models were created over both targets using as a primary datasource multi-mission Ku band altimetry and other satellite data. Using an augmented version of the method used to identify salars, criteria were established to identify and mask river pixels. A further distinction was made to identify wetland / seasonally inundated regions, and detailed masks were produced for areas to exclude from soil moisture work. Comparing the Congo basin DREAM and its mask with independent data (Dargie et al., 2017) revealed the wealth of surface hydrology information encoded in the beta DREAM model. For the Congo beta test DREAM, 13% of the DREAM pixels are identified as river surfaces and 34% as wetland/seasonally flooded areas. It is noted that many smaller tributaries are below the current spatial resolution of the DREAM, and are classified with their surrounding terrain as wetland pixels. For the Amazon beta test DREAM, the corresponding statistics are 23% rivers and 36% wetlands.
These figures show the proportion of the models masked from soil moisture determination. Over what proportion of this surface are data retrieved by Ku band altimeters? To determine this, the masks were tested with multi-mission altimeter data. A waveform analysis system was utilised to assess echo shapes, scan for complex waveforms and flag echoes from water surfaces. Waveform shapes are classified using a system which identifies fourteen classes of echo shape corresponding to known surface types. The system is tuned for each instrument and observing mode using calibration areas of known characteristics. Multi-mission statistics show highest data retrieval over rivers and wetlands, lower over unmasked DREAM pixels. This is an expected outcome, as excluding rivers and wetlands selects for rougher topography. Varying proportions of waveforms were flagged by the system as returns affected by pools of still water throughout the model areas, with the highest proportions from the Amazon basin.
Backscatter data from all instruments show excellent agreement with the DREAMs, with cross-correlation coefficients with data from dry terrain better than 0.9. Altimeter soil moisture datasets are shown to demonstrate good agreement with external validation data. Small soil moisture signals are successfully recovered from desert regions, where other techniques encounter difficulties.
The ability of nadir-pointing altimeters to penetrate vegetation canopy gives a unique perspective in rainforest areas. Over the Amazon and Congo basins, the DREAM masking process creates detailed maps of river and wetland extents, with over 60% of the Amazon and 50% of the Congo DREAM areas identified as rivers, wetlands and seasonally flooded regions. The clear implication is that, to monitor surface water optimally in these rainforests (within the constraints of satellite orbit and repeat period), satellite altimeters should retrieve data from the majority of the underlying surface. Fortunately, analysis of past altimeter performance shows that this goal was largely achieved for the Congo and Amazon basins, particularly by ERS2 and EnviSat. Waveform analysis is found to be essential to exclude returns affected by pools of water within the altimeter footprint. Surface soil moisture time series can then be derived, and are shown to correlate with adjacent river height time series.
Very limited data acquisition from Sentinel-3A, due to the current OLTC mask, critically constrains the scope of SRAL DREAMing over all DREAMs, but results are consistent both with Cryo-Sat2 SAR and LRM mode data and results from prior missions.
In conclusion, satellite radar altimetry can provide soil surface moisture estimates wherever a DREAM can be crafted. Altimeter soil moisture estimates contribute to the datastore over river basins, providing an independent assessment of soil moisture data from other sources.
Waveform classification and soil moisture retrieval works for SRAL altimeters, with good results from Sentinel-3A, where data are available.
Data are currently being analysed to craft DREAMs over further river systems.

References
Dorigo, W.A., Wagner, W., Albergel, C., Albrecht, F.,  Balsamo, G., Brocca, L., Chung, D., Ertl, M., Forkel, M., Gruber, A., Haas, E., Hamer, D. P. Hirschi, M., Ikonen, J., De Jeu, R. Kidd, R.  Lahoz, W., Liu, Y.Y., Miralles, D., Lecomte, P. (2017).  ESA CCI Soil Moisture for improved Earth system understanding: State-of-the art and future directions. In Remote Sensing of Environment, 2017,  ISSN 0034-4257, https://doi.org/10.1016/j.rse.2017.07.001.
Dargie, GC; Lewis, SL; Lawson, IT; Mitchard, ET; Page, SE; Bocko, YE; Ifo, SA; (2017) Age, extent and carbon storage of the central Congo Basin peatland complex. Nature , 542 pp. 86-90. 10.1038/nature21048.

As the severity and occurence of flood events tend to intensify worldwide with climate change, the need for high fidelity flood forecasting capability increases. However, this capability remains limited due to a large number of uncertainties in models and observed data. In this regard, the Flood Detection, Alert and rapid Mapping (FloodDAM) project, funded by Space Climate Observatory (SCO) initiatives, was set out to develop pre-operational tools dedicated to enabling quick responses in selected flood-prone areas, as well as improving the resolution, reactivity and predictive capability of existing decision support systems.

Hydraulic numerical models are used in hindcast mode to improve knowledge on flood dynamics, assess flood-related damage and design flood protection infrastructures. They are also used in forecast mode by civil security agencies in charge of decision support systems, for flood monitoring, alert, and management. These numerical models are developed to simulate and predict water surface elevation (WSE) and velocity with lead times ranging from a couple of hours to several days. For instance, Telemac2D (www.opentelemac.org) solves the Shallow Water Equations with an explicit first-order time integration scheme, a finite element scheme and an iterative conjugate gradient method. However, such models remain imperfect because of the uncertainties in their inputs that translate into uncertainties in the model outputs. These uncertainties are related, for instance, to the simplified equations, the numerical solver, the forcing and boundary conditions or to the model parameters resulting from batch calibration, such as friction and boundary conditions.

Data Assimilation (DA) allows to reduce these uncertainties by sequentially combining the numerical model outputs with observations, as they become available, and taking into account their respective uncertainties. These techniques are widely used in geosciences and have proven to be effective in river hydrodynamics and flood forecasting. The Ensemble Kalman Filter (EnKF) is implemented here to reduce uncertainties in upstream time-varying inflow discharge to the river catchment as well as in spatially distributed friction coefficients, with the assimilation of in-situ WSE data at observing stations. The optimality of the EnKF depends on the ensemble size over which covariances are stochastically estimated and on the observing network, especially in terms of its spatial and temporal density. The use of remote-sensing (RS) data allows to overcome the limits due to the lack and decline of in-situ river gauge stations, especially in flood plains. In recent years, Synthetic Aperture Radar (SAR) data has been widely used for operational flood management due to the ability to map flood extents over large areas in near real time, and its all-weather day-and-night image acquisition capabilities. Water bodies and flooded areas typically exhibit low backscatter intensity on SAR images, since most of the radar pulses are, upon arrival at the water surfaces, specularly reflected away. Therefore, these areas can be detected relatively straightforward from SAR images, with exceptions in built-up environments and vegetated areas. In the present work, RS-derived flood extents are obtained by a Random Forest (RF) algorithm applied on Sentinel-1 images. The RF was trained on a database that gathers manually delineated flood maps from Copernicus Emergency Management Service Rapid mapping products from past flood events. It also takes into account the MERIT DEM to improve the flood detection precision and recall.

This work highlights the merits of assimilating RS-derived flood extents along with in-situ data that are usually confined in the river bed, in order to improve the representation of the flood plain dynamics. Here, the RS-derived flood extents are post-processed to express the number of wet and dry pixels over selected regions-of-interest (ROI) in the floodplain. These pixel-count observations are assimilated along with in-situ WSE observations to account for errors in friction and upstream forcing. They provide spatially distributed information of the river and flood plain but with a limited temporal resolution that depends on the satellite overpass times; for instance Sentinel-1 has a revisit frequency of several days (maximum six days) while in-situ observations are available every 5 to 15 minutes for observing stations in the VigiCrue network (https://www.vigicrues.gouv.fr/).

The study area is the Garonne Marmandaise catchment (south-west of France) which extends over a 50-km reach of the river Garonne, between Tonneins and La Réole. The control vector for the EnKF-DA is composed of seven friction coefficient values (six on the main channel and one for the floodplain) and three corrective parameters to the inflow discharge. Results are shown for a flood event that occurred in January-February 2021, with a forecast lead time up to +24 hours. It was shown that the assimilation of both RS and in-situ data outperforms the assimilation of in-situ data only, especially in terms of 2D dynamics in the flood plains. Quantitative performance assessments have been carried out by comparing the simulated and observed water level time-series at in-situ observing stations and by computing 2D metrics computed between the simulated flood extent maps and the SAR-derived maps (i.e. Critical Success Index and F1-score based on the expression of the contingency table). This work paves the way toward a cost-effective and reliable solution for flood forecasting and flood risk assessment over poorly gauged catchments or even ungauged catchments. Once generalized, such developments could potentially lead to hydrology-related disaster risk mitigation in other regions. Future progresses built upon this work will extend to other catchments and the assimilation of other flood observations.

For more than two decades, satellite altimetry has demonstrated the potential to derive water level time series of inland waters. Nowadays, accuracies of water level time series between few centimeters for large lakes and few decimeters for smaller lakes and rivers can be achieved. However, there is still potential for quality improvements when optimizing the processing strategy, for example in view of retracking algorithms, off-nadir effects, or outlier rejection.

In 2015, DGFI-TUM published the first DAHITI approach that is based on an extended outlier rejection and a Kalman filter approach. In this poster, we present an updated DAHITI approach, which considers the following aspects for deriving high-accurate water level time series for small inland waters: First, detailed analysis of the altimeter sub-waveforms is performed in order to detect that part of the radar echo that can be assigned to the water bodies of interest. Additionally, off-nadir reflections are analyzed and taken into account, in order to derive reliable error information of the water level time series. This step is also the first step of the outlier rejection, which is extended by applying other criteria. For example, this additionally contains a detection of ice coverage. In order to achieve long-term consistent and homogenous water level time series, the latest geophysical corrections and models are applied and a multi-mission crossover analysis is performed for all altimeter missions.

We present preliminary results for selected inland waters, which are validated by using in-situ data. The results of the new DAHITI approach show a significant improvement of the accuracy of water level time series and its errors estimation.

Climate change increases the likelihood of catastrophic flood events, resulting in destruction of cropland and infrastructure, thereby threatening food security and exacerbating epidemics. These dangerous impacts highlight the need for rapid monitoring of inundation, which is necessary to estimate the dimensions of the disaster. An accurate satellite-based flood mapping can support the risk management cycle, from near-real-time rescue and response until post-event analysis. Current remote sensing techniques allow cheap, quick, and accurate flood classifications, using freely accessible satellite-data, for instance, from the Copernicus Sentinel satellites. Indeed, the Synthetic Aperture Radar (SAR) sensor on-board Sentinel-1 (S1), is uniquely suited to flood mapping due to its 24-hour weather independent imaging technology, and is widely used globally due to the open data availability. Binary classifications are widely used to extract flood inundation from SAR data, but due to the large discrepancy in prevalence of flood/non-flood classes in an S1 tile, finding adequate appropriate labelled samples to train classifiers is extremely challenging in addition to being time consuming. Furthermore, the process of training data collection is non-trivial due to a variety of uncertainties in SAR data originating from the underlying land-use and incorrect labeling could lead to gross misclassifications. For example, if the training data does not sufficiently represent the flood surface roughness diversity, large inundated tracts could be missed by the classifier. Consequently, training a binary can be expensive, slow, and compromise on accuracy, since precise labels for both classes are required despite only one class of interest.

One-class classifiers address this issue, by using only samples of the class of interest, i.e. the true positives, making them the perfect choice for flood classification. Even though one-class classifiers have outperformed classical binary classifiers for a variety of use-cases, surprisingly they have not been widely used so far in flood mapping literature. Accordingly, this study provides the first assessment of one-class classifiers for flood extent delineation from SAR data.

The study area is the coastal part of Beira, Mozambique, where the Cyclone Idai made landfall on 15th March 2019. Idai was the deadliest cyclone in the Southern hemisphere, affecting over 850.000 people and leading to a Cholera outbreak. S1 SAR data was used to classify the inundated area using Support Vector Machine (SVM) and Random Forest (RF) for the binary classification and one-class SVM (OCSVM) for the one-class classification. The data inputs and training data for both flood classifications were the same. For validation concurrent cloud-free Sentinel-2 (S2) optical-data were used.

Preliminary results suggest that one-class classifiers can perform equivalently or better than standard classifiers for flood detection from SAR images given similar volume of training data. Moreover, one-class classifiers offer the advantage of using limited training data and thus result in lower classifier training as well as processing time, without compromising on detection accuracy. Based on the results obtained in this first benchmarking study, the use of one-class classifiers for flood mapping should be further explored, for a robust performance assessment given different underlying land-uses and geographical regions.

The changes in the runoff and in the alluvial outflow lead to changes in the slope, the depth, meandering, the width of the river bottom and the vegetation. The bed load and the suspended load can change the morphology of the river bed as a result of high runoff. This has a direct impact on the determination of the fairway in navigable rivers. That is why it is of great importance for assisting the maintenance of the navigable rivers to provide with instruments to predict the modifications in the river morphology that will potentially impact the fairway. Achieving this has also effect for understanding of the freshwater cycle, for developing our knowledge of the Earth. To address this problem it is necessary to forecast the sediment deposition amounts and the river runoff and to determine how they will change the river morphology. Predicting sediment deposition potential depends on a variety of meteorological and environmental factors like turbidity, surface reflectance, precipitations, snow cover, soil moisture, vegetation index. Satellite data offer rich variety of datasets, supplying this information. We adopt deep learning to address some specifics of Earth observation data, such as their inconsistency, and generate missing data in the time-series with generative adversarial networks - GANs. And consequently we apply the rendered consistent earth observation data along with in-situ measurements on other deep learning architectures (convolutional neural networks CNNS and LSTMs) to actually generate forecasts for river runoff, water level and sediment deposition by using historic satellite data of the meteorological features listed above, and in-situ measurements for water level, runoff and turbidity. Thus, we employ earth observation data for developing AI based solutions that translates as EO4AI. Further, we report on a series of prediction models and experiments carried out on data from the downstream of the Danube and from Arda that show precision of forecasts with minimal deviation with respect to real measurements. To leverage the applicability of the forecasts on the river morphology in integrated models, we calibrate hydrodynamic models using Telemac, and we demonstrate how the fusion of a complex EO4AI method and geometry mapping produces a solution for a real user need of being aware of upcoming changes in the river fairway of the downstream of the Danube. The satellite data are provided by ADAM via the NoR service of ESA. ADAM provides data access to satellite datasets from different satellites with semantic relevance for the construction of sediment transport and deposition forecast model as discussed above. Finally, we demonstrate a visualization of the forecasted fairway on a GIS component using ESRI ArcGIS server.

Acknowledgement
This work has been carried out within ESA Contract No 4000133836/21/NL/SC

Monitoring the water resources from Space is a rapidly developing area of application for radar altimetry. Recent progress in instrumentation (development of SAR and SARIn altimeter sensors) and radar signal treatment has allowed us for the first time to include medium and small continental water objects in the scope of application of altimetry. In the Republic of Ireland only the lakes with area more than 10 km2 are included in the in situ water level observational network. This represents only 30% of the total lake area. The island dimensions (~84 400 km2) limit development of long fluvial networks. As a result, the width of even the largest river channels does not exceed 250 m and ranges mainly within 20 - 100 m. The potential of radar altimetry for monitoring lakes of 2-4 km2 area and rivers of 80-200 m width has already been demonstrated in prior studies. We explore the capacity of the most recent generation of satellite altimetry missions (Jason-3, CryoSat-2, and Sentinel-3) for monitoring water bodies, water courses and water regime of peatlands on the entire territory of the Republic of Ireland. In the framework of the ESA HYDROCOASTAL Project we 1) investigated the performance of SAR (Sentinel-3) and conventional (Jason-2,-3) altimetry to retrieve water level time series, 2) evaluated the advantage of enhanced (80Hz) Sentinel-3 sampling rate processing provided by the ESA G-POD/SARvatore online and on-demand SAR Altimetry Processing Service and 3) examined the gain from the combination of measurements of repeat-orbit (Sentinel-3) and geodetic-orbit (CryoSat-2) satellite missions. We also investigated an effect of river width and configuration of the fluvial virtual station on the accuracy of the water height retrievals as well as assessed the impacts of the surrounding relief on the performance of satellites to produce high quality altimetric water level time series that may be of value to a broad variety of users.

At high latitudes, ice cover on lakes and rivers is the key factor of local and regional climatic, environmental and socioeconomic systems. It modulates heat and mass exchange with the atmosphere, reshapes riparian ecosystems, and may induce hazardous flooding. In many remote regions of the Arctic, freshwater (river and lake) ice is a crucial actor for socioeconomic resilience of local communities. It provides: 1) a unique infrastructure for the transport of goods and people via winter ice roads; 2) access to fishing and hunting grounds; and 3) supplies drinking water. Each year hundreds of kilometres of roads are built in Canada, Alaska, Russia, Norway, Finland, and Switzerland on lake ice and river ice by regional/local authorities or by local residents. For safe usage of ice roads, a variety of information on ice parameters (initial freeze-up, structure, thickness and growth history, fracturing, metamorphism, etc.) is needed. During the last few years, several European (ESA CCI+ Lakes, ESA LIAM, CNES TOSCA) and Russian (RFBR "Arctic") projects have funded research dedicated to investigation of freshwater ice from space. In this presentation, we provide several examples of the use of satellite observations for the study of lake and river ice parameters and discuss results in context of their potential application for safe ice cover use on Lake Baikal and on the Ob River (Siberia).

On Lake Baikal, intra-thermocline eddies often form prior to ice formation and continue to develop under the ice cover. These eddies weaken and melt the ice. Several areas of frequent eddy appearance are located in sections of the lake where ice roads are used by local people and not monitored operationally. The combination of the different optical, imaging SAR and radar altimetry missions helps to monitor and understand the spatial distribution of eddies and the transformation of ice cover by their presence. On the Ob River, radar altimetry observations were used for retrieving ice phenology dates and ice thickness along a 400-km river reach. The retrievals demonstrated a good potential for the forecasting of the ice road operation in Salekhard City. In situ observations are needed for adequate interpretation of satellite observations in the context of changing ice properties. Radiative transfer modelling can also be helpful and, in the near future, may allow for the estimation of the main freshwater ice parameter of interest - ice thickness. Here, we present the first results of the application of the Snow Microwave Radiative Transfer (SMRT) model for the simulation of radar altimeter backscatter and emissivity of Lake Baikal ice during winter 2018-2019.

Satellite remote sensing is an effective approach to monitor floods over large areas. Ground-based gauges remain a vital instrument for monitoring water levels or streamflow, but they cannot capture the spatial extent of a water body or flood. Numerical models can be an excellent source of such information, but are not readily available in all regions and can be costly to set up. Satellites already orbit and monitor nearly all regions of the globe and can thus provide relevant information where other sources are lacking. However, while earth observation has many advantages, there are also data gaps and challenges, which can be different for each specific sensor.

Flood mapping studies and applications often use imagery from optical, e.g. MODIS, Landsat, Sentinel-2, and/or synthetic aperture radar (SAR) sensors, e.g. ALOS, Sentinel-1. SAR’s cloud penetrating capability is especially important for flood mapping, as clouds are often present over (inland) floods, because these are triggered by rainfall originating from clouds. ESA’s Sentinel-1 constellation has for the first time in history made it possible to provide reliable flood mapping services on a large (even global) scale. The synergistic use of optical imagery can help overcome some of SAR’s known issues regarding flood mapping (such as signals resembling that of water over sandy soils and/or agricultural fallows), as well as help provide more timely flood maps, essential for disaster response and relief efforts. Still, current satellite-derived flood maps are not perfect and under- and overestimations of flood waters are to be expected. This is especially true for areas under thick vegetation canopies, as both optical and (most) SAR sensors cannot penetrate these, and urban areas, where signals can be distorted and data from the freely available satellites mentioned here don’t possess the spatial resolution required to accurately map water between or within urban features.

The HYDrologic Remote sensing Analysis for Floods (HYDRAFloods) tool is already using multiple sensors for the improved capabilities mentioned above, with current research focusing on data fusion of optical and SAR imagery as well as the inclusion of hydrologic information. Hydrology plays an important role in the general water cycle, influences floods and can also be used to constrain or improve satellite-derived flood maps. Low soil moisture values in sandy soils and areas of agricultural fallows can be used to prevent false positives derived from SAR imagery. Hydrologically-relevant topography information can be used in a similar fashion, but also to identify potentially flooded areas that are otherwise obscured from satellite imagery, such as under forest canopies. For this, we link the flood maps to hydrologically connected surface water flow paths.

HYDRAFloods is under active development in the SERVIR-Mekong program, covering a large part of Southeast Asia, by ADPC, SIG, SEI and Deltares, supported by NASA and USAID. It is used operationally by the United Nations World Food Programme (WFP) in Cambodia, being made available in their Platform for Real-time Impact and Situation Monitoring (PRISM), and was field tested during the severe floods that hit the country in October 2020. HYDRAFloods embraces open science and combines relevant algorithms from literature with our own custom developments, which are published in open access journals. It runs on the Google Earth Engine platform to facilitate easy data access and running at scale across the entire South East Asia region. The code itself is hosted on an online repository with open source license, including up-to-date documentation.

HYDRAFloods has been described in general at other conferences, so we will only give a brief overview and instead focus on recent research on including hydrologically relevant information in the processing chain to obtain more accurate flood maps. We hope this can lead to a fruitful discussion on the underlying techniques and assumptions, as well as contribute to a broader discussion on combining data from various sources (e.g. in-situ, models, EO) and its best practices.

The Sentinel-6 mission, launched in November 2020, carries the first radar altimeter operating in open burst with a PRF high enough (~9kHz) to perform the focussing of the whole target observation echoes in a fully coherent way with practically negligible impact from along-track replicas. Furthermore, such a feature allows improvement to the along-track resolution down to the theoretical limit around 0.5 m when processing the data with a Fully-Focussed SAR (FFSAR) algorithm. This resolution increment actually represents a revolutionary step with respect to the ~300 m along-track resolution provided by current operational processors based on Unfocussed SAR algorithms, commonly used in previous radar altimeters with a closed burst chronogram, such as CryoSat-2 and Sentinel-3. In this contribution, we explore new applications over inland water surfaces derived from such new Sentinel-6 FFSAR products. Indeed, the FFSAR Ground Prototype Processor (GPP), developed by isardSAT and based on the backprojection algorithm [1], has been used to process data over different types of inland water targets with the following objectives: (1) validate range measurements with in-situ water height data in case of nadir targets and (2) monitor water extent for off-nadir targets located within certain observation constraints. As a main outcome, we present a methodology to estimate water extent of small targets located on unambiguous across-track targets. We have analysed targets that present strong seasonal variability in terms of area, and validated the method by comparing water extent measurements derived from Sentinel-6 with the ones derived from optical, SAR imagery and in-situ observations. The overall work is part of the VALERIA (Validating Algorithms Levels 1A and 2 in Ebre RIver Area) project developed within the Sentinel-6 Validation Team using data from the satellite commissioning phase.

[1] Egido, Alejandro and Walter H. F. Smith. “Fully Focused SAR Altimetry: Theory and Applications.” IEEE Transactions on Geoscience and Remote Sensing 55 (2017): 392-406.

Global hydrological models simulate water storages and fluxes of the water cycle, which is important for e.g. water management decisions and drought/flood predictions. However, models are plagued by uncertainties due to the model input errors (e.g. climate forcing data), model parameters, and model structure resulting in disagreements with observations. To reduce these uncertainties, models are often calibrated against in-situ streamflow observations or compared against total water storage anomalies (TWSA) derived from the Gravity Recovery And Climate Experiment (GRACE) satellite mission. In recent years, TWSA data are integrated into some models via data assimilation.

In this study, we present our framework for jointly assimilating satellite and in-situ observations into the WaterGAP Global Hydrological Model (WGHM). For the first time, we assimilate three data sets:
(a) GRACE-derived TWSA,
(b) in-situ streamflow observations from gauge stations; this is in preparation for the Surface Water and Ocean Topography (SWOT) satellite, which will be launched in 2022 and is expected to allow the derivation of streamflow observations globally for rivers wider than 50-100m, and
(c) Global SnowPack snow coverage data derived from the Moderate Resolution Imaging Spectroradimeter (MODIS), which is installed on NASA’s Earth Observing System satellites.

GRACE assimilation strongly improves the TWSA simulations within the Mississippi River Basin, e.g. the correlation increases to 91%, with which our results are consistent with previous studies. However, we find in this case that the streamflow simulation deteriorates, for example, correlation reduces from 92% to 61% at the most downstream gauge station. In contrast, jointly assimilating GRACE data and streamflow observations from GRDC gauge stations improves the streamflow observations by up to 33% in terms of e.g. RMSE and correlation while maintaining the good TWSA simulations. We use the snow coverage data first to independently validate the impact of TWSA and streamflow assimilation on the snow simulation, and then, for the first time, assimilate the snow coverage data into the WGHM. We expect that this will not only further enhance the streamflow simulations but also the simulations of single WGHM water storages like the snow storage.

Water volumes available in natural and artificial lakes are of prime interest, either for water management purposes or water cycle understanding. However, less than 1% of global lakes are monitored. To such an end, remote sensing has been a useful tool providing continuous and global information for more than 30 years.

Combining information of water elevation with altimetry along with water surface from optical and SAR images can lead to the relative volume variation through the creation of a height-surface-volume relationship (hypsometric curve). This method is currently limited by the altimetry data coverage which is non global (less than 3% worldwide).

Even if the future wide swath altimeter, SWOT, will provide the first global survey of water bodies, the estimation of bathymetry and corresponding hypsometric curve remains a challenge to estimate water volume. Contextual approach can be considered and even trained to approximate a reservoir bathymetry from a “filled” DEM. We used such a contextual approach to develop an algorithm using deep learning to recreate the reservoir’s bathymetry.

The first step consisted in using Digital Elevation Models (DEM) cropped to the sub-basin provided by the Hydrobasin shapefile database. This led to the creation of an artificial database of DEM patches with their associated pseudo-water basins and associated 20m high reservoir. This approach was applied to relatively “dry” but mountainous or hilly countries like Chile, Turkey, Morocco, among others. A specific attention has been given to avoid planar DEM area from already existing water dams with varying water heights from 5 to 20meters. We also checked for each sub-basin that the created virtual reservoir was realistic, for instance in terms of dam length or related water surface. Thereby, we created around 9000 DEM/water surfaces patches to train a Unet deep learning algorithm. We also used data augmentation, data refining and cross-validation over the simulated reservoirs to get a realistic model. The recreated bathymetry led to an error lower than 10% on volume estimation and still improving at this time.

Further advances could be applied not only to reservoirs, but also to lakes and rivers. This would improve the global water volume estimations, but also discharges, thanks to the ever-improving DEM datasets produced in terms of precision and resolution, like with the future CO3D mission (CNES).

Flooding is one of the most damaging natural hazards, causing economic losses and threats to livelihoods and human health. Predicting flooding patterns in river systems is a challenge in large and remote areas. Climate change has intensified the occurrence of severe flood events, and accurate predictive models are required for flood risk management. However, the available in-situ data in remote areas and ungauged river basins is scarce and often not available in the public domain. The accuracy of hydrodynamic models is limited by the quality of the available observations, which are essential to calibrate unobserved or unobservable model parameters.

Accurate topographic elevation measurements are essential to replicate 1D/2D flood processes. Satellite based DEMs have the advantage of providing large coverage in remote areas. However, high resolution DEMs are not always available and therefore it is necessary to use lower resolution products. Missions such as Shuttle Radar Topography Mission (SRTM) or ALOS-PALSAR offer DEMs down to 1-Arcsec resolution freely available. When used as input to hydraulic models, such DEMs can lead to large errors in simulated water surface elevation and surface water extent, due to the complex topography of floodplains that is normally hard to map precisely. To better integrate this data in the model, a finer resolution product will be needed to map the floodplain topography and river bathymetry. The novel altimetry mission ICEsat-2, operating since 2018, offers a large spatial coverage, with an along track resolution down to 70 cm in its photon cloud product ATL03. This data has shown great potential when mapping river topography, identifying narrow river structures and also when multi channel rivers and braided structures are present. ATL03 can be used as a control point dataset, to correct the biases and refine DEMs.

In this study we use a 1D hydraulic model derived from ICEsat-2 data to characterize the river bed geometry of the main channel. We use ATL03 product to map the topography of the river channel, providing accurate data on river bed geometry. We calibrate depth and manning roughness against ATL13 water surface elevation observations, which is the inland water product from ICEsat-2. The water surface elevation is simulated with an accuracy to decimeter level providing a precise river bathymetry characterization. To study the water surface elevation in the floodplain, we combine the SRTM DEM 1-Arcsec resolution with ATL03 cross-sections, to reduce elevation errors. With the refined DEM, we run Mike Flood 1D/2D inundation module, and validate the simulated inundation areas with flood maps from Global Surface Water Explorer.

The developed workflow is demonstrated for sections of Amur river, which flows in China and Russia. This river is characterized by large floodplains and for having a braided structure, making it a suitable study case to demonstrate our methodology.

The Mediterranean regions are characterized by intense rainfall events strongly affected by violent rainfall events causing floods. The vulnerability to flooding in the Moroccan High Atlas, especially in the Tensift basin, has been increasing over the last decades. Rainfall-runoff models can be very useful for flash flood forecasting. However, event-based models require a reduction of their uncertainties related to the estimation of initial moisture conditions before a flood event. Soil moisture may strongly modulate the magnitude of floods and is thus a critical parameter to be considered in flood modeling.
The aim of this study is to compare daily soil moisture measurements obtained by time domain reflectometry (TDR) at Sidi Rahal station with satellite soil moisture products (European Space Agency Climate Change Initiative, ESA-CCI), in order to estimate the initial soil moisture conditions for each event. A modeling approach based on rainfall-runoff observations of 30 sample flood events from (2011 to 2018), in the Ghdat basin, were extracted and modeled by an event-based rainfall-runoff model (HEC-HMS) which is based on the Soil Conservation Service (SCS-CN), loss model, and a Clark unit hydrograph was developed for simulation and calibration of the 10-minute rainfall runoff.
These data were used in the validation process of the event modeling part and indicate that soil moisture could help to improve the initial conditions of event-based models in small basins to improve the quality of flood forecasting. The rationale is that a better representation of the catchment states leads to a better streamflow estimation. By exploiting the strong physical connection between the soil moisture dynamic and rainfall, this methodology is very satisfactory for reproducing rainfall-runoff events in this small Mediterranean mountainous watershed, since Nash coefficients of validation are ranging from (0.76 to 0.89), the same approach could be implemented in other watersheds in this region. The results of this study indicate that the remote sensing data are theoretically useful for estimating soil moisture conditions in data-sparse watersheds in arid Mediterranean regions.
Keywords: Soil moisture; Floods; Remote sensing; Hydrological modeling, CN method, Mediterranean basin.

In semi-arid regions and especially in Sahel, water bodies such as small reservoirs, small lakes, and ponds are vital resources for people. Most studies on inland waters in Africa focus on large lakes like Lake Chad for example, but the numerous lakes and ponds, which are found near almost every village in Sahel, are poorly known. These small water bodies (SWB) are critical in terms of water resources and important for greenhouse gases and biodiversity. SWB probably increased in numbers and surface recently, due to changes in land surface properties after the big Sahelian drought of the late 20ieth century (Gal et al. 2017, doi 10.5194/hess-21-4591-2017), and to dam building, as for instance in Burkina Faso. For a more detailed assessment of changes in water resources, it is necessary to quantify water volumes variability and hydrological regime of these SWB at the regional scale.

The objectives of this work are to develop methods to monitor water quantity of SWB by combining optical and radar remote sensing. This study is carried out over 3 countries (Niger, Mali and Burkina Faso) and addresses the water regime of 40 water bodies over the 2016-2021 period.

Water surface is derived from Sentinel2 optical data. Algorithms for water detection generally face two issues in this region: i) the high number of vegetated water bodies (floating vegetation, grasses or trees), ii) the extremely high and unusual reflectance of Sahelian waters. It turned out that a threshold on the MNDWI index chosen ad hoc for each lake, implemented in Google Earth Engine, is a fast and efficient method to estimate water areas. Water levels are derived from Sentinel 3 altimetry data processed with the ALTIS software (Frappart et al. 2021, doi 10.3390/rs13112196). Careful extraction is required for water bodies in close proximity, such as the Tanvi reservoirs in Burkina Faso, since multiple signals coming from neighbouring water bodies may mix in the radar data.

Water levels and matching water areas are combined to derive surface-height curves. This allows estimating water levels from any Sentinel2 data, and densifies therefore the water levels time-serie derived from Sentinel3 altimetric data alone. Time-serie of water levels are then used to estimate water levels decrease during the dry season (generally from November to June), which is compared to the evaporation loss from each SWB estimated using Penman's method and ERA5 data.

Given that during the dry season water inflow by precipitation is null, differences between water level and evaporation are due to water losses, or uptakes from anthropogenic activities or exchanges with groundwater or river networks. For the 40 SWB studied, evaporation averages about 7 mm/d during the dry season, whereas water losses vary significantly across different water bodies. Water bodies exposed to important pumping activities exhibit significantly higher water loss rates, than the evaporation rate, thus reaching a minimum water balance value around –12.5 mm/d. Others water bodies display the opposite situation, for example for lakes in the inner Niger Delta where the flood extends into the dry season and water is supplied by groundwater or river network, with water balance around 5.7 mm/d.

The results show the potential of the water balance approach in poorly observed semi-arid regions to better understand hydrological processes, including human management of reservoirs. This is particularly relevant for the forthcoming SWOT mission, which will enable this approach to be applied at the global scale.

Continental and global hydrological models are the primary means to simulate surface/sub-surface water storage, water flux, and surface water inundation variables, which are required for hazard mitigation and policy support plans. However, establishing these large-scale models is challenging since the complicated physical processes that govern large-scale hydrology cannot be fully resolved by the simplified equations in these schemes. Besides, it is well known that the model parameters are insufficient to account for intensification of the water cycle caused by the climate change and anthropogenic modifications. Another issue is that most hydrological and hydraulic models are at best only calibrated against river discharge or similar data, but these calibrated parameters may have limited influence on the estimation of water storage and water volume changes in large-scale basins. In this study, we demonstrate the extent to which Terrestrial Water Storage (TWS; a vertical summation of surface and sub-surface water storage) data from the Gravity Recovery And Climate Experiment (GRACE) and its follow-on mission (GRACE-FO), as well as remotely sensed soil moisture data can improve the estimations of river discharge and water extent as well as water storage during episodic droughts and floods. For this, we present the structure of our in-house ensemble Kalman filter based calibration and data assimilation (C/DA) as well Bayesian model-data merging frameworks to integrate freely available satellite data into the in-house modified W3RA water balance model forced by ERA5 data. The results are demonstrated through simulations of water storage, river discharge, drought characteristics and floods in West Africa and Europe.

The number of active gauges with open-data policy for discharge monitoring along rivers has decreased over the last decades. Therefore, we cannot properly answer crucial questions about the amount of freshwater available on a certain river basin, the spatial and temporal dynamics of freshwater resources, or the distribution of the world’s freshwater resources in the future. The recent breakthroughs in spaceborne geodetic techniques enable us to overcome the lack of comprehensive measurements of freshwater resources and allow us to understand the hydrological water cycle more realistically. Among different techniques for estimating river discharge from space, developing a rating curve between the ground-based discharge and spaceborne river water level or width is the most straightforward one. However, this does not always lead to promising results, since the power law rating curves describe a river section with a regular geometry. Such an assumption may cause a large modeling error. Moreover, rating curves do not deliver a proper estimation of discharge uncertainty as a result of the mismodelling and the coarse assumptions made for the uncertainty of inputs.

Here, we propose a nonparametric model for estimating river discharge and its uncertainty from spaceborne river width measurements. The model employs a stochastic quantile mapping function scheme by, iteratively: 1) generating realizations of river discharge and width time series using Monte Carlo simulation, 2) obtaining a collection of quantile mapping functions by matching all possible permutations of simulated river discharge and width quantile functions, 3) adjusting the measurement uncertainties according to the point cloud scatter. The algorithm’s estimates are improved in each iteration by updating the measurement uncertainties according to the difference between the measured and estimated values.

We validate the proposed algorithm over 14 river reaches along the Niger, Congo, Po and Mississippi rivers. Our results show that the proposed algorithm can mitigate the effect of measurement noise and also possible mismodelling. Moreover, the proposed algorithm delivers a meaningful discharge uncertainty. Evaluating the discharge estimates via the stochastic nonparametric quantile mapping function and the rating curve technique shows that the performance of the proposed algorithm is superior to the rating curve technique especially in challenging cases.

With an along-track resolution of around 300 m, ESA CryoSat-2 (CS2) brought along a whole new range of monitoring possibilities of inland water bodies. The introduction of Synthetic Aperture Radar (SAR) altimetry enabled the study of rivers and lakes that were not visible with conventional Low Resolution Mode (LRM) altimeters. However, the 300 m resolution is still a challenge for the smallest water bodies, for which sometimes none or only a single observation is available.
Over some selected water bodies, the CS2 altimeter operates in SAR Interferometric (SARIn) mode, using both the antennas on board. The phase difference between the two returns can be used to locate the across-track origin of the echo. While, traditionally, retracking methods are used to retrieve a single surface height estimate from waveforms over inland water bodies, in this study, we apply a swath approach where multiple peaks of single SARIn waveforms are retracked and geolocated across track using the phase difference information.
We show that this method can be used to retrieve a large number of valid water level estimates (WLE) for each SARIn waveform, even from water bodies that are not immediately located at the satellite nadir. We investigate the potential of this technique over rivers and lakes by looking at the increase in spatial coverage as well as at the impact on the precision of the measurements when compared with conventional nadir altimetry and in-situ hydrometric data.
Increasing the number of WLE is of great importance especially for small water bodies, where the number of available valid measurements from altimeters is generally very limited. The results presented in this work are additionally relevant for the future Copernicus Polar Ice and Snow Topography Altimeter mission (CRISTAL), which will also fly an interferometric altimeter.

Restricted access to freshwater and crop failure lead to disastrous consequences, for example economic losses, hunger and death. Thus, ensuring food production and sufficient water supply for crop production (or agriculture) is a highly relevant topic for the population all over the world. Soil moisture is the main driver for providing water resources for agriculture and vegetation but in semi-arid or arid regions it is becoming more important to derive water from surface water bodies or stored in groundwater. These surface and subsurface water storages are either monitored with in-situ data, which have a long record history, or are simulated in models, which provide global simulations with a good spatial resolution (~50 km). However, the in-situ data are not spatially explicit and very sparse and, thus, cannot cover each climate regime and the models encounter problems with uncertainty in the forcing data and model assumptions.

In the last decades, the use of remote sensed data has enabled observation of the water from space. GRACE (Gravity Recovery And Climate Experiment) and its successor GRACE-FO were and are so far the only satellite missions that observe the sum of surface and subsurface water with global resolution. However, GRACE(-FO) have a coarse spatial resolution (~300 km) and only sense the vertically aggregated sum, so called total water storage anomalies (TWSA); hence a further separation into the different water compartments is needed. Therefore, we integrate GRACE into a hydrological model via assimilation to improve the model’s realism while spatially downscaling and vertically disaggregating GRACE.

In this study, we assess signatures and subsignals found in models using observation (via assimilation) based storages and vegetation (via remote sensing) measures derived from MODIS (Moderate Resolution Imaging Spectroradiometer). In a case study, we interrogate two main processes (measured at peak times) in South Africa for 2003 to 2016: 1) The precipitation-storage dynamics, i.e. the dynamics of the pathway from precipitation to replenished soil moisture, surface water and groundwater and 2) the storage-vegetation dynamics, i.e. the pathway from the corresponding storage to vegetation growth (by using in Leaf Area Index and Actual Evapotranspiration).

Generally, we found that the amount of water that refills the storages is often overestimate in the modeling and the duration for this process is often shorter compared to the observations. For example, we found in the modeling that in general the annual peak of groundwater lags the annual precipitation peak by 3 months, while the observations identify a 4-month lag. For the storage-vegetation dynamics we also notice an overestimation of the amount of water that contributes to vegetation growth, and an over- or underestimation of the duration for this process strongly depends on the considered storage. Our study concluded that the model did not correctly capture the precipitation-water storages-vegetation dynamics and it would be impossible to conclude that from using only GRACE TWSA data, without data assimilation. In the future, our findings will be highly relevant for modelers to and can be used to improve the model structures.

Agricultural systems are the main consumers of freshwater resources at global scale, using 60 % to 90 % of the total available water. While the growing demand for agricultural products and the resulting intensification of their production will increase the dependency on available freshwater resources, this sector will become even more vulnerable because of the intensifying impacts of climate change. Detailed knowledge about soil moisture, being a key parameter in the agricultural sector, can help to mitigate these effects. Nevertheless, spatial and temporal high resolution surface soil moisture data for regional and local monitoring (down to precision farming level) are still challenging to obtain. By using current as well as future Synthetic Aperture Radar (SAR) satellite missions (e.g. Sentinel-1, ALOS-2, NISAR, ROSE-L), this knowledge gap can be filled. Providing a cloud- and weather independent monitoring of the Earth's surface, SAR observations are suitable for regional and local soil moisture estimations, but with a global extent. While the increasing resolution and total number of SAR recordings will contribute to an improvement of the estimation in general, the computational costs as well as the local memory capacity on the other hand become a limiting factor in processing the vast load of data. Here, on-demand cloud-based processing services are one way to overcome this challenge. This is especially interesting as most of the severely affected regions have limited access to computational resources.
Using both VV and VH polarization for vegetational detrending as well as low pass filtering, we developed an automated workflow for estimating soil moisture using temporal and spatial high-resolution Sentinel-1 timeseries, based on the alpha approximation approach of Balenzano et al. 2011. The workflow is established within the cloud processing platform Google Earth Engine (GEE), providing a fast and applicable way for on-demand computation of soil moisture for individual time periods and areas of interest around the globe. The algorithm was tested and validated over the Rur catchment, located in the federal state of North-Rhine Westphalia in the West of Germany. With an area of 2,354 km², it comprises a great diversity in agricultural cropping structure as well as topologies. A total of 711 individual Sentinel-1A and Sentinel-1B dual-polarized (VV + VH) scenes in Interferometric Wide-Swath Mode (IW) and Ground Range Detected High Resolution (GRDH) format are used for the analysis from January 2018 to December 2020. Using all available orbits (both ascending and descending), a temporal resolution of one to two days could be achieved with a spatial resolution of 200 m. The workflow includes multiple steps: speckle filtering, incidence angle normalization, vegetational detrending and low-pass filtering. The results were validated against eight Cosmic-Ray Neutron Stations (CRNS), which are evenly distributed over the catchment, covering various types of landcover. In total, the method achieves an unbiased RMSE (uRMSE) of 5.84 % with an R² of 0.46. Looking at individual months, the highest correlation can be achieved in the months April and October with R² values range between 0.65 to 0.7, while the lowest correlation is observed in July and January, with R² values ranging between 0.15 and 0.2. Looking at individual landuse, the method achieves the best results for pastures, with an uRMSE of 0.42 and an R² value of 0.63.

Estonia is known for its large riverside areas that are seasonally (in spring) flooded over. However, extremely warm winters in Estonia during the last five years have also caused large floodings during the winter. Changes in inundation extent, depth, and duration can change the phonological patterns, animal migration routes and affect the forest management, resulting in economic losses. Therefore, a need to assess the inter-annual variability of inundation along riverside areas has become interest from both public and private sectors.
At the European scale, two flood-monitoring services are provided: The (1) Copernicus Emergency Management Service provides a free-of-charge mapping service in cases of natural disasters, man-made emergencies, and humanitarian crises throughout the world. This service can be triggered by request in the case of an emergency. The (2) Copernicus Land Monitoring Service provides a pan-European high-resolution product, Water and Wetness. This product shows the occurrence of water and wet surfaces over the 2015-2018 period.
However, these services cannot be used for the inter-annual identification of flooded areas. Therefore, an automatic processing scheme of Sentinel-1 data was set up for the mapping of open-water flood (OWF) and flood under vegetation (FUV). The methodology was applied for water mapping from Sentinel-1 (S1) and a flood extent analysis of the three largest floodplains in Estonia in 2019/2020. The extremely mild winter of 2019/2020 resulted in several large floods at floodplains that were detected from S1 imagery with the maximal OWF extent up to 5000 ha and maximal FUV extent up to 4500 ha. A significant correlation (r2 > 0.6) between OWF extent and closest gauge data was obtained for inland riverbank floodplains. The outcome enabled us to define the critical water level at which water exceeds the shoreline and flooding starts. However, for a coastal river delta floodplain, a lower correlation (r2 < 0.34) with gauge data was obtained and the excess of river coastline could not be related to a certain water level. At inland riverbank floodplains, the extent of FUV was three times larger compared to that of OWF. The correlation between the water level and FUV was < 0.51, indicating that the river water level at these test sites can be used as a proxy for forest floods.
The analysis of the extent and frequency of wintertime floods can form the basis for various economic analyses as well as evaluations of revenue being conducted in the forest industry due to mild winters and evaluations of stress to Northern boreal alluvial meadows. Relating conventional gauge data to S1 time series contributes to the implementation of flood risk assessment and management directives in Estonia

Monitoring water levels can help hydrological modeling, predict hydrological responses to climatic and anthropogenic changes, and ultimately contribute to environmental protection and restoration. However, measuring lake water levels is easier said than done. The conventional ground-based gauges are now scarce due to limited accessibility, high cost, the labor needed for continuous maintenance, and required security and oversight of equipment. Although satellite altimetry is a standard tool for water level change detection in lakes worldwide, the newest sensors still have limitations regarding coarse temporal and spatial resolution and re-tracking errors from the backscattered signal from non-water surfaces. Changes in water levels can also be retrieved from Differential Interferometric Synthetic Aperture Radar (DInSAR) by measuring the phase change between two Satellite Radar images, but these changes are relative in space and difficult to unwrap.
Here, we develop a new methodology to estimate the absolute water level changes in the only 30 small Northern-latitude lakes gauged in Sweden. Sweden has more than 100,000 lakes covering 9% of the country’s surface area. We aim to evaluate the capability of InSAR in estimating absolute water level changes of lakes in latitudes beyond 55 degrees without the need of unwrapping the phase component, as it is usually done for InSAR studies over water surfaces. With the constraint of a very short temporal baseline (6 days) between pair Sentinel-1 SAR images, we deal with the phase jump in interferograms resulting from sudden changes in water level, and instead of unwrapping each interferogram, we accumulate the phase change of successive pair images across nine months in 2019. We chose only pixels inside the lakes’ surface area that exhibit a steady, coherent behavior across all interferograms and identified the pixels where the DInSAR and gauged estimates of water level change show high linear correlation coefficients (R2 > 0.8). We found lakes with many pixels showing a high correlation, suggesting the capability of DInSAR to determine the direction of water level change in these lakes. The highest correlation between the accumulated phase change and the gauged water level was observed in a pixel on Lake Båven, southeast Sweden (R2 > 0.97), and the lowest correlation was observed in a pixel on Lake Lillglän in the west of the country (R2 > 0.26). The pixels with a high correlation between the accumulated phase change and the gauged water level were located along the lake shorelines, surrounded by forest and wetland land covers. Surprisingly, features on these shores can still enable the double bounce of the radar signal necessary for the interferometric technique, allowing the retrieval of water level change. The high correlation in these pixels shows that the accumulated phase change of the Sentinel-1 twin satellites can help detect the trends of water level change in high latitude lakes surrounded by marsh-dominated wetlands and forests or other shoreline features.

The SAR-mode processing in altimetry as it is currently operated in ground segments does not
exploit the full capabilities of the SAR system in terms of spatial resolution. The so-called unfocused
SAR altimeter (UFSAR) processing performs the coherent summation of pulses over a limited number
of successive pulses (64-pulses bursts of a few milliseconds in length) reducing the along-track resolution
down to 300 m only. Since recently [2], the concept of coherent summation has been extended
to the whole illumination time of the surface (typically more than 2 seconds) allowing to increase
the along-track resolution up to the theoretical limit (approximately 0.5 m) and thus improving the
SAR-mode capability for imaging reflective surfaces of small size. The benefits of the fully-focused
coherent processing have already been demonstrated on various surfaces to differentiate targets of
heterogeneous surfaces (like sea-ice, inland-water and coastal) and to achieve the maximum effective
number of looks available from SAR altimetry on homogeneous surfaces (like ocean) [2].
The limitations of the FF-SAR in closed-burst mode has already been reported in [2], creating very
harmful artificial side lobes in the along-track dimension due to lacunary chronogram. It is extremely
challenging to separate real signal from its replicas when they are superimposed, considering that
every reflecting focalization point on ground creates its own replicas. Both Sentinel-3 and Cryosat-2
SAR altimetry missions have been designed with a lacunary chronogram, one exception is for the
quasi-continuous pulse transmission Sentinel-6 inter-leaved mode. Over heterogeneous targets, replicas
interference creates peaks and troughs pattern, with overflow of power outside of the water body
boundaries and destruction of power inside the water body boundaries. This clearly jeopardizes confidence
in data and its use for large water body detection like lead detection a major goal in SAR
altimerty sea-ice application.

At level-2, impact of replicas on the estimated geophysical parameters are not completely understood
yet. Even at crossing point between Sentinel-3 and Sentinel-6, it might be tricky to compare the
results due to not identical footprint and overflight angle, but also altimeters differences apart from
chronogram (like the sampling frequency, deramping/match-filtering and SNR). A new methodology
of comparison has been developed and implemented at CLS taking Sentinel-6 data and emulating the
sparse closed-burst mode chronogram of Sentinel-3 by removing pulses. Thus on same acquisition
points open-burst and closed-burst can be compared each other by isolating only replica effect. More
than 700 hydrological targets (including narrow rivers, larger rivers, lakes and dam) have been already
processed. First results showed as expected global differences in amplitude but more surprisingly a
higher range variability of 1.5cm in closed-burst mode compared to open-burst mode.

Next progress is replica removal, a very important topic if we expect to exploit the full potential
of FF-SAR processing with Sentinel-3 and Cryosat-2 data. We propose a deconvolution technique to
recover the open-burst radargram using optimization method starting from Wiener filtered first guess
[3] and a model that takes into account replicas. A new model of multi-scatterer FF-SAR impulse response
function, based on LRM inland-water model approach in [1], has been developed and validated
over diverse rivers data acquisition. This model supposes water presence a priori knowledge, which
might be relevant for inland-water (by exploiting water surface masks), but turns out to be completely
irrelevant for sea-ice lead targets permanently in movement. To tackle this problem, the replica model
is first optimized to determine the position and specularity of water presence that fit at best real data.
Once the model fixed, deconvolution is validated by comparison of reference Sentinel-6 open-burst
geophysical parameters with deconvoluted degraded Sentinel-6 closed-burst data. Different surfaces
captured by Sentinel-6 will be deconvoluted including rivers, lakes, leads and open-ocean.

Keywords— FFSAR, closed-burst, replica, deconvolution

References
[1] R. Abileah, A. Scozzari, and S. Vignudelli. Envisat RA-2 Individual Echoes: A Unique Dataset for a Better
Understanding of Inland Water Altimetry Potentialities. Remote Sensing, 9(6):605, June 2017.
[2] A. Egido and W. H. F. Smith. Fully Focused SAR Altimetry: Theory and Applications. IEEE Transactions
on Geoscience and Remote Sensing, 55(1):392–406, Jan. 2017.
[3] A. Monti-Guarnieri. Adaptive removal of azimuth ambiguities in SAR images. Geoscience and Remote Sensing,
IEEE Transactions on, 43:625–633, Apr. 2005.

Fluvial and riparian ecosystems have many important ecological, social, and economic functions. Several EO-based tools and products have been developed for their monitoring at a global scale. However, the spatial resolution of these global-level products is often too coarse for monitoring narrow rivers and especially their upper sections. These are the areas where rivers are most dynamic and where frequent and accurate monitoring is particularly pressing. To overcome spatial resolution limitations, we developed a method for river monitoring using fraction maps produced with linear spectral signal unmixing.

We developed and tested the method on the Soča and Sava rivers in Slovenia and the Vjosa river in Albania. We mapped three land cover classes of interest – surface water, vegetation, and gravel. The use of spectral bands in combination with the NDVI, MSAVI2, NDWI, and MNDWI indices produced the best results. We achieved similar accuracies with endmembers selected manually and endmembers selected automatically with the N-FINDR algorithm. The optimal total number of endmembers used for spectral signal mixture analysis was found to be between three and five. A larger number of endmembers led to clustering of spectral signatures and thus redundant information. Tests showed that the inclusion of shade as a separate endmember did not improve fraction map accuracy. Furthermore, we found that endmembers selected manually or automatically on one satellite image can be successfully transferred to analyse another image acquired in a comparable geographic region and at a similar phenophase.

Results of the soft classification were compared to hard classification using the Spectral Angle Mapper with the same endmembers. Fraction maps were more accurate than maps based on hard classification both for Sentinel-2 and Landsat images. Water presence detected on the fraction maps was correlated with in situ measured water level and river discharge with Pearson’s r > 0.6 (p < 0.0001). We examined the ability of fraction maps to detect changes in river morphology. By looking at three different timestamps (13 October 2017, 3 July 2019, 5 September 2020), the results showed that fraction map differencing could distinguish changes in gravel deposition down to 400 m2 in extent. We found that change detection accuracy was best on pixel level when changes amounted to at least 30%. Finally, we tested the possibility of detecting river morphology changes from a time series of land cover presence based on fraction maps. The extents of water and gravel can vary following changes in water level. However, we found that a decrease of gravel bar size within two standard deviations of the mean indicated regular variations while a larger decrease pointed to gravel bar removal.

The developed method can be used for monitoring fluvial and riparian environments in highly heterogeneous areas. The main limitations of the method are associated with cloud obstruction and terrain shadow that are known problems of optical images. An interesting line of future investigations is to test the possible contribution of using SAR data for fluvial morphology monitoring.

Title: Quality Flag and Uncertainties over Inland Waters

Inland waters are essential for environmental, societal and economic services such as transport, drinking water, agriculture or power generation. But inland waters are also one of the most affected resources by climate change and human populations growth.
Altimetry, which has been used since 1992 for oceanography, has also proven to be a useful tool to estimate inland water surfaces such a rivers and lakes, which are considered Essential Climate Variables (ECVs). the heterogeneity of the targets size, surfaces roughness etc… and the surrounding environment near the water targets make the interpretation of the measurements more complex. In addition, the availability of a measurement must be complemented by the confidence that it can be attributed to the estimation of the water surface height and providing the uncertainty associated with this measurement will be useful for assimilations and downstream products.
The aim of this presentation is to describe the use of a waveform classification method, based on neural network algorithms, on level 2 data in order to identify reliable measurements on water body targets. This classification can be used as a metric for data quality and is therefore incorporated in the data processing to define a quality flag in the inland product. The quality flag is being implemented in two ESA projects using data for the reprocessing of several missions data: FDR4ALTwith data from ENVISAT, ERS-2 and ERS-A missions and CryoTempo with data from Cryosat2. Secondly, it aims at the presenting the methodology for estimating the uncertainty on the estimated water level.

References:
Birkett C. M. (1995). Contribution of TOPEX/POSEIDON to the global monitoring of climatically sensitive lakes, Journal of Geophysical Research, 100, C12, 25, 179-25, 204
J.F. Cretaux, W. Jelinski, S. Calmant, A. Kouraev, V. Vuglinski, M. Berge-Nguyen, M.C. Gennero, F. Nino, R.A. Del Rio, A. Cazenave, P. Maisongrande: SOLS: A lake database to monitor in the Near Real Time water level and storage variation from remote sensing data, Advances in Space Research, N47, ELSEVIER, 2011,pp. 1497-1507
Crétaux J-F and C. Birkett, lake studies from satellite altimetry, C R Geoscience, Vol 338, 14-15, 1098-1112, doi: 10.1016/J.cre.2006.08.002, 2006
Poisson, Jean-Christophe & Quartly, Graham & Kurekin, Andrey & Thibaut, Pierre & Hoang, Duc & Nencioli, Francesco. (2018). Development of an ENVISAT Altimetry Processor Providing Sea Level Continuity Between Open Ocean and Arctic Leads. IEEE Transactions on Geoscience and Remote Sensing. PP. 1-21. 10.1109/TGRS.2018.2813061.
N. Longépé et al., "Comparative Evaluation of Sea Ice Lead Detection Based on SAR Imagery and Altimeter Data," in IEEE Transactions on Geoscience and Remote Sensing, vol. 57, no. 6, pp. 4050-4061, June 2019, doi: 10.1109/TGRS.2018.2889519.

The seasonal snow cover in mountains is crucial for ecosystems and human activities. Developing methods to map snow depth at high resolution ("< 10 m") is an active field of snow studies as snow depth is a key variable for water ressource and avalanche risk assessment. Most methods rely on close range remote sensing, combining lidar or photogrammetry with an airplane or a drone. However, drone acquisitions are limited to small areas ("< 10 km²") and airborne campaigns are logistically difficult to set up in many mountains of the world. Satellite photogrammetry is an innovative method for monitoring the seasonal snowpack in mountains and could help address the challenge of estimating the distribution of snow in any place of the world. Accurate snow depth maps at high spatial resolution ("~ 3 m") are calculated by differencing digital elevation models with and without snow derived from satellite stereoscopic images.
Here we present a collection of snow depth maps calculated from 50 cm Pléiades stereoscopic images in the central Andes, the Alps, the Pyrenees, the Sierra Nevada (USA) and Svalbard. The comparison with a reference snow-depth map measured with airborne lidar in the Sierra Nevada, provides a robust estimation of the Pléiades snow depth error. At the 3 m pixel scale, the standard error is about 0.7 m. The error decreases to 0.3 m when the snow-depth maps are averaged over areas greater than "10^3 m²". Specific challenge arose in some sites due to the lack of snow free terrain or due to artefacts inherent to satellite images. However, Pléiades snow depth maps are sufficiently accurate to allow the observation of snow redistribution patterns due to wind transport and avalanche, or the precise determination of the snow volume in a "100 km²" catchment. Assimilated in a distributed snowpack model, Pléiades snow depth amps improve the modeled spatial variability of the snow depth and compensate for lacking processes in the model or bias in the meteorological forcings. The available collection of Pléiades snow depth maps provides the opportunity to characterize with a consistent method the snow cover in an unprecedented variety of sites, such as the arctic, alpine mountains and subtropical regions.

Flooding is the most frequent natural hazard on Earth and affects an increasing number of people. Major events are responsible for huge loss of life and substantial destruction of infrastructure. Detailed information about the location, time, or extent of present and historic floods help in improving emergency response or planning of prevention actions. For this purpose, the new Global Flood Monitoring (GFM, https://gfm.portal.geoville.com) service provides satellite-based flood mapping information derived from Sentinel-1 Synthetic Aperture Radar (SAR) data in near-real time (NRT) on a global scale to the user community (Salamon et al, 2021). This service is part of the Copernicus Emergency Management Service (CEMS), and is available in its beta-version through the Global Flood Awareness System (GloFas, https://www.globalfloods.eu/). In order to improve the overall reliability of the flood mapping, three independent Sentinel-1-based algorithms are combined within one ensemble product.

As basis for all activities within the GFM service, a global Sentinel-1 datacube has been created (Wagner et al, 2021). In the initial phase, more than 1.6 million Sentinel-1 scenes from 2015 – 2020 were preprocessed using the new 30m Copernicus DEM for terrain correction. The observations were resampled to a spatial gridding of 20m and are provided in a tiled and stacked image structure based on the Equi7Grid (https://github.com/TUW-GEO/Equi7Grid).This setup allows for an efficient extraction of spatiotemporal subsets. The Sentinel-1 datacube is updated in NRT to enable continuous flood monitoring.

One of the algorithms going into the ensemble product is the algorithm developed by the Technische Universität Wien (TU Wien, https://www.geo.tuwien.ac.at/). The algorithm performs a pixel-wise decision between flooded and non-flooded conditions. The historic Sentinel-1 measurements of the datacube and derived temporal parameters allow to statistically describe the backscatter signature of both states. The water backscatter differs significantly from non-flooded land due to the specular reflection of the impinging radiation and the side-looking geometry of the SAR system. Contrary to water surfaces, the backscatter signals over non-flooded land are much more heterogeneous and most show strong seasonal variations. This seasonality is caused by variable factors within the signal like soil moisture or vegetation conditions. To parametrise the backscatter under non-flooded conditions, and by considering the backscatter’s seasonality, a harmonic regression model was found to be best suited (in particular for NRT operations). The model’s parameters were computed for each pixel of the Sentinel-1 datacube by a least-square estimation which made use of measurements from 2019-2020. Based on the resulting global parameter database and the underlying model, one is able to estimate the non-flooded backscatter for every day of the year. Using Bayes interference, the incoming Sentinel-1 scene is pixel-wise compared to the modelled backscatter signature of flooded and non-flooded conditions, and the more probable condition is then chosen.

When working with SAR data, water-look-alikes like deserts, radar shadows, or tarmacs could be confused easily with inundated areas. Additionally, one is limited to areas where the Sentinel-1 signal is able to reach the ground undisturbed in order to distinguish between flooded and non-flooded situations. Consequently, areas which are densely vegetated or built-up areas need to be excluded as well as areas which permanently feature low backscatter. Therefore, we utilise exclusion layers that are derived from temporal parameters of the Sentinel-1 datacube. By masking the flood mapping results with the exclusion layers, potential uncertainties are avoided and the algorithm’s robustness is increased.

In this contribution, we present the TU Wien Sentinel-1 flood mapping algorithm, which exploits the historic measurements of a dedicated Sentinel-1 2015-2020 datacube, and which is already integrated within the GFM ensemble approach. We evaluate the globally operated algorithm in representative sites of a set of world regions, highlighting its strengths and caveats. Additionally, we focus on the suitability of the Sentinel-1 signal history to exclude areas that show low sensitivity for flood mapping or could potentially be classified wrongly as flooded.

Salamon et al. (2021) The New, Systematic Global Flood Monitoring Product of the Copernicus Emergency Management Service. In 2021 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), IEEE, pp. 1053-1056.

Wagner, Wolfgang, et al. (2021) A Sentinel-1 Backscatter Datacube for Global Land Monitoring Applications. Remote Sensing 13.22 , 4622.

Earth Observations (EO) have become popular in hydrology because they provide information in locations where direct measurements are either unavailable or prohibitively expensive to make. Recent scientific advances have enabled the assimilation of EO’s into hydrological models to improve the estimation of initial states and fluxes which can further lead to improved forecasting of different variables. When assimilated, the data exert additional controls on the quality of the forecasts; it is hence important to apportion the effects according to model forcings and the assimilated data. Here, we investigate the hydrological response and seasonal predictions over the snow-melt driven Umeälven catchment in northern Sweden. The HYPE hydrological model is driven by two meteorological forcing datasets: (i) a down-scaled GCM product based on the bias-adjusted ECMWF SEAS5 seasonal forecasts, and (ii) historical meteorological data based on the Ensemble Streamflow Prediction (ESP) technique. Six datasets are assimilated consisting of four EO products (fractional snow cover, snow water equivalent, and the actual and potential evapotranspiration) and two in-situ measurements (discharge and reservoir inflow). We finally assess the impacts of the meteorological forcing data and the assimilated data on the quality of streamflow and reservoir inflow seasonal forecasting skill for the period 2001-2015. The results show that all assimilations generally improve the skill but the improvements vary depending on the season and assimilated data. The lead times until when the data assimilation influences the forecast quality are also different for different datasets and seasons; as an example, the impact from assimilating snow water equivalent persists for more than 20 weeks during the spring. We finally show that the assimilated datasets exert more control on the forecasting skill than the meteorological forcing data, highlighting the importance of initial hydrological conditions for this snow-dominated river system.

In the last couple of decades, active remote sensing technologies, such as radar or LiDAR based sensors, became an essential source of information for the monitoring of inland water body levels. This is due to their validated high accuracies [1], and as a way to fill-in for the ever-decreasing water-level gauge stations that is reported worldwide [2,3].
In this study, we are interested in evaluating the accuracy, and correcting, water level estimates from the recently launched Global Ecosystem Dynamics Investigation (GEDI) full waveform (FW) LiDAR sensor on board the International Space Station (ISS). GEDI, which became operational in 2019, is equipped with three 1064 nm lasers with a pulse repetition frequency (PRF) of 242 Hz. One of the lasers’ power is split in two while the remaining two operate at full power. These four lasers are equipped with beam dithering units (BDUs) that rapidly deflect the light by 1.5 mrads in order to produce eight tracks of data. The acquired footprints along the eight tracks are separated by 600 m across track, and 60 m along the track, with a footprint diameter of 25 m.
Since the launch of GEDI, there have been few studies that assessed its accuracy for the estimation of in-land water levels [4–6]. The first study conducted by Fayad et al. [4], used the first two months of GEDI acquisitions (mid-April to mid-June 2019) to assess the accuracy of GEDI altimetry over eight lakes in Switzerland. For these two months, they reported a mean difference between GEDI and in situ gauge water elevations (bias) ranging from -13.8 cm (under estimation) to +9.8 cm (over estimation) with a standard deviation (SD) of the bias ranging from 14.5 to 31.6 cm. The study conducted by Xiang et al. [6] over the five great lakes of north America (Superior, Michigan, Huron, Erie and Ontario) using five months of GEDI acquisitions (April to August 2019) found a bias ranging from -32 cm (under estimation) to 11 cm (over estimation) with a SD that ranged from 15 to 34 cm. Finally in the study of Frappart et al. [5] which assessed the accuracy of GEDI data over ten Swiss lakes using acquisitions spread over seven months (April to October 2019) found a bias that ranged from -15 cm (under estimation) to +21 cm (over estimation) with a SD ranging from 10 cm to 30 cm.
The factors influencing the physical shape of the waveform and therefore the accuracy of LiDAR’s altemetric capabilities can be grouped into three categories: (1) Instrumental factors (e.g. viewing angle, signal over noise ratio), (2) water surface variations factors (e.g. wave heights and period, wave type), and (3) Atmospheric factors (e.g. cloud presence and cloud composition). For example, the viewing angle at acquisition time was demonstrated to increase elevation errors for ICESat-1 GLAS when the viewing angle deviates from nadir due to precision attitude determination [7]. Water specular reflection is also another potential source of errors due to the saturation of the detector [8]. Finally, clouds and their composition are major factors that affect the quality of LiDAR acquisitions [9,10]. Indeed, and while opaque clouds attenuates the LiDAR signal thus the receiver only captures noise, less opaque clouds allow the LiDAR to make a full round trip, but could potentially increase the photon path length due to forward scattering (atmospheric path delay), thus resulting in biases in elevation measurements [11]. Moreover, GEDI’s return signal strength will greatly vary between cloud-free shots and clouded acquisitions [9].
The objective of this study is therefore two folds. First, the performance of GEDI’s altimetric capabilities using filtered (i.e. removal of noisy acquisitions) GEDI waveforms across the five great lakes (Lakes Erie, Huron, Ontario, Michigan, and Superior) was assessed. Next, a random forest regression model was trained in order to estimate the calculated difference between GEDI acquisitions and in situ water level records using the instrumental, water surface variations and atmospheric variables as predictors to this model. The output of this model, which is namely the estimated difference between each GEDI acquisition and its corresponding in situ reference, was subtracted from each GEDI’s acquisition elevation in order to produce corrected elevation estimates.
Results showed that uncorrected GEDI estimated have on average a bias of 0.3 m (ranged between 0.25 and 0.42 m) and a root mean squared error (RMSE) of 0.58 m (ranged between 0.54 and 0.67 m). After the application of our model, the bias was mostly eliminated (ranged between -0.07 and 0.01 m), and the average RMSE decreased to 0.17 m (ranged between 0.14 and 0.21 m).

References

1. Birkett, C.; Reynolds, C.; Beckley, B.; Doorn, B. From Research to Operations: The USDA Global Reservoir and Lake Monitor. In Coastal Altimetry; Vignudelli, S., Kostianoy, A.G., Cipollini, P., Benveniste, J., Eds.; Springer Berlin Heidelberg: Berlin, Heidelberg, 2011; pp. 19–50 ISBN 978-3-642-12795-3.
2. Shiklomanov, A.I.; Lammers, R.B.; Vörösmarty, C.J. Widespread Decline in Hydrological Monitoring Threatens Pan-Arctic Research. Eos Trans. AGU 2002, 83, 13, doi:10.1029/2002EO000007.
3. Hannah, D.M.; Demuth, S.; van Lanen, H.A.J.; Looser, U.; Prudhomme, C.; Rees, G.; Stahl, K.; Tallaksen, L.M. Large-Scale River Flow Archives: Importance, Current Status and Future Needs. Hydrol. Process. 2011, 25, 1191–1200, doi:10.1002/hyp.7794.
4. Fayad, I.; Baghdadi, N.; Bailly, J.S.; Frappart, F.; Zribi, M. Analysis of GEDI Elevation Data Accuracy for Inland Waterbodies Altimetry. Remote Sensing 2020, 12, 2714, doi:10.3390/rs12172714.
5. Frappart, F.; Blarel, F.; Fayad, I.; Bergé-Nguyen, M.; Crétaux, J.-F.; Shu, S.; Schregenberger, J.; Baghdadi, N. Evaluation of the Performances of Radar and Lidar Altimetry Missions for Water Level Retrievals in Mountainous Environment: The Case of the Swiss Lakes. Remote Sensing 2021, 13, 2196, doi:10.3390/rs13112196.
6. Xiang, J.; Li, H.; Zhao, J.; Cai, X.; Li, P. Inland Water Level Measurement from Spaceborne Laser Altimetry: Validation and Comparison of Three Missions over the Great Lakes and Lower Mississippi River. Journal of Hydrology 2021, 597, 126312, doi:10.1016/j.jhydrol.2021.126312.
7. Urban, T.J.; Schutz, B.E.; Neuenschwander, A.L. A Survey of ICESat Coastal Altimetry Applications: Continental Coast, Open Ocean Island, and Inland River. Terrestrial Atmospheric and Oceanic Sciences 2008, 19, 1–19.
8. Lehner, B.; Döll, P. Development and Validation of a Global Database of Lakes, Reservoirs and Wetlands. Journal of hydrology 2004, 296, 1–22.
9. Fayad, I.; Baghdadi, N.; Riedi, J. Quality Assessment of Acquired GEDI Waveforms: Case Study over France, Tunisia and French Guiana. Remote Sensing 2021, 13, 3144, doi:10.3390/rs13163144.
10. Shu, S.; Liu, H.; Frappart, F.; Kang, E.L.; Yang, B.; Xu, M.; Huang, Y.; Wu, B.; Yu, B.; Wang, S.; et al. Improving Satellite Waveform Altimetry Measurements With a Probabilistic Relaxation Algorithm. IEEE Trans. Geosci. Remote Sensing 2021, 59, 4733–4748, doi:10.1109/TGRS.2020.3010184.
11. Yuekui Yang; Marshak, A.; Palm, S.P.; Varnai, T.; Wiscombe, W.J. Cloud Impact on Surface Altimetry From a Spaceborne 532-Nm Micropulse Photon-Counting Lidar: System Modeling for Cloudy and Clear Atmospheres. IEEE Trans. Geosci. Remote Sensing 2011, 49, 4910–4919, doi:10.1109/TGRS.2011.2153860.

This study developed a method to derive field-specific SM information (as opposed to the large-footprint existing products) in near-real time by leveraging synergies of hydrological models and Earth observation (EO) data, both from SAR and optical sensors. The two components are further complemented by EO near-real time information on meteorological fields for drivers of precipitation and evapotranspiration. While the strength of the soil hydrological models consists of a physically based description of the rain infiltration and percolation processes, the satellite-based data permits to derive vegetation canopy properties at the field scale, and to obtain forcing variables such as precipitation and potential evapotranspiration to feed the models.
For several fields in the COSMOS UK soil moisture monitoring network, we retrieved time series of Sentinel-2 NDVI and Sentinel-1 backscattering values. We used the Hydrus-1D modelling tool for the simulation of surface and in-depth SM at the study fields at daily time steps during periods of low vegetation (NDVI < 0.25). The model’s upper boundary conditions were given by the time series of satellite-based estimates of precipitation and evapotranspiration. The lower boundary condition was set as free drainage, assuming that the water table is deeper than the root zone and the soil is well drained.
For C-band SAR and for the Sentinel-1 range of incidence angles, the literature reports an approximate linear relationship between the average VV backscattering coefficient of uniform, bare-soil crop fields and the surface soil moisture. For individual fields during the bare soil stage, it can be assumed that the main cause of Sentinel-1 VV backscattering changes is surface moisture, more rapidly variable than the surface roughness. Then, the fields’ soil hydraulic conductivity and other infiltration descriptors were obtained by optimizing the temporal trends of the modelled surface moisture to the temporal trends observed in the Sentinel-1 VV backscattering during low vegetation periods.
The soil moisture simulated in this way was compared to the moisture measured at the COSMOS fields at different depths. Our method achieved an excellent accuracy, only drifting away from the measured values at the end of the cropping cycle, after harvesting.
This work was carried out in collaboration with Mantle Labs Ltd. And received funding from the UK Research and Innovation SPace Research & Innovation Network for Technology (SPRINT) programme. By deriving valuable soil moisture information at the field level, Mantle Labs intends to offer enhanced drought related insurance products which can be made available to smallholder farmers. This index insurance will protect farmers against crop loss occurring due to extreme weather events.

One of the less understood feedbacks is the role of burrowing animals on the soil hydrology. Burrowing animals were shown to increase soil macroporosity and affect vegetation distribution, both of which have huge impacts on infiltration, preferential flow, surface runoff, water storage and field capacity. However, the specific role of burrowing animals on these variables is to date poorly understood, the presence of burrowing animals has largely not been included in the erosion models. A suitable approach which enables studying their impacts on the catchment scale and compare them across several climate zones is missing but needed to fully understand the feedbacks between the pedo- and biosphere.

To close this research gap, we combined in-situ measurements of soil properties and burrow distribution, high resolution remote sensing data and machine-learning methods with numerical modelling.

For this, we first conducted field surveys on the presence and absence of animal burrows along a predefined track with 8 the hillside. We extracted 160 soil samples along the catena of study hillsides, as well as 316 soil samples from animal burrow areas and control areas without an animal burrow. We analysed them on several physical and chemical properties needed for model parametrisation and as well estimated the difference between samples extracted from burrow area and control area. We studied the daily surface processes at the burrow scale and measured the volume of excavated sediment by the animals and the sediment redistribution processes within the burrow area during rainfall events using laser scanners for a period of 7 months.

Then, we combined the in-situ measured soil properties and the burrow distribution with remote sensing and machine learning and upscaled the soil properties and the presence of animal burrows to each catchment at a resolution of 0.5 m. We conducted a land cover classification to estimate the vegetation cover and combined LiDAR data with the DGM to estimate the vegetation height.

We implemented the upscaled soil properties, burrow locations and vegetation parameters into Morgan-Morgan-Finney model and parametrised one model per catchment. For this, we adjusted the input parameters at the burrow locations according to the measured soil properties, vegetation cover and height, estimated microtopography changes, burrowing behaviour and sediment excavation and redistribution within the burrow area. We validated the model using the in-situ installed sediment traps.

To estimate the impacts of burrowing animals, we ran the model with included and not included animal burrows. We estimated the daily and yearly impacts of the presence of the burrows on soil erosion, infiltration, preferential flow, surface runoff, water storage and field capacity.


We present a parametrised model, which includes the presence of animal burrows in its calculation and the modelled impacts of burrowing animals on soil erosion, infiltration, preferential flow, surface runoff, water storage and field capacity on the catchment scale at a 0.5 m resolution. We compare the short-term and long-term impacts on the soil hydrological properties at the burrow and catchment scale along the climate gradient.


The numerical model achieved an accuracy of R2 = 0.70. The presence of burrows had a positive impact on sediment erosion, infiltration and water storage and negative impact on surface runoff and field capacity. These effects were on the daily and burrow scale most pronounced in the semi-arid and mediterranean climate zone. In the semi-arid climate zone, the burrows heavily affected already sparse vegetation which then affected the surface infiltration and runoff. In the mediterranean climate zone, the burrow size and entrance diameter especially had an impact on the preferential flow. At the catchment and yearly scale, the effects were most pronounced in the humid zone. Although the density of burrows was here low, due to regularly occurred rainfall events the burrowing animals here cumulatively contributed the most to all hydrological processes. In the arid zone, the impact of burrowing animals was detectable during sporadically occurring heavy rains.

Our study thus shows the potential of inclusion of burrowing animals into numerical models as well as the importance to do so, as our results show heavy impacts of the presence of the burrows on hydrological processes in all climate zones at various temporal and spatial scales.

The inland water monitoring is crucial to estimate the volume of flow into the channel and quantify the available water resource to supply the human needs. This has an essential role for both society and environment and due to the numerous issues related to the ground hydro-monitoring networks it represents a political and economic challenge. A valuable alternative to derive the surface water information on a global scale involves satellite earth observations and over the last decades, the satellite altimetry has proven to be a well-established method for providing water level measurements.
In the context of the ESA-funded FDR4ALT (Fundamental Data Records for Altimetry) project, innovative Earth system data records called Fundamental Data Records are used to produce the Inland water Thematic Data Record based on the exploitation of measurements acquired by the altimeter onboard ERS-1, ERS-2 and ENVISAT.
In this work, we present the first results of the project, showing the analysis of the different retrackers (Ice1, Ice3, MLE4, TFMRA, Adaptive) on different water bodies, such as rivers and lakes of different sizes and environments during different periods of the year. A Round Robin analysis is carried out to evaluate the performances of each retracker with the final goal to detect the best retracker able to describe the inland water flow to be implemented at global level. The performances are calculated against external measurements from reliable ground-based measurements and using datasets from other altimetry sources freely available on the web (Theia Hydroweb, Dahiti, HydroSat).

Rivers play an important role in regulating and distributing inland water resources in the processes of the hydrological cycle of the earth, which is an important factor for the steady development of regional economy and climate change. The river width, water level and flow velocity are important parameters to characterize the changes in river discharge. With the rapid development of remote sensing technology and hydrological models, the width, velocity and slope of rivers can be effectively estimated. But the monitoring river water level in high-precision is not effective, especially for small and medium-sized river basins, due to the low spatial and temporal resolution and the lack of measurement precision. We develop a new retracker to process the inland water altimetry waveforms, called AMPDR-PF (Automatic Multiscale-based Peak Detection Retracker using Physically-based model Fitting). We compare the water level estimated by AMPDR-PF with water level from official altimeter products in the River Rhine. We finally use it to estimate river discharge in the Rhine river.

The mix of quantitative and qualitative methods (AMPDR-PF) are considered to retrack the inland water altimetry waveforms for improving the accuracy of the river level at different spatial scales. Point of departure are combining the advantages of the AMPDR method and SAMOSA+ methods. Moreover, the new method allow for sensitivity analysis in different altimeter data such as Sentinel-3A/3B and Sentinel-6, accuracy validation such as the standard deviations of overpasses, RMSEs (the root-mean-squared errors) compared with the tide gauge at different spatial scales. Time-series of Water Surface Elevation (WSE) from multiple virtual stations are built after correcting for the river mean slope. Additionally, time-series of river width and river slope are generated from Sentinel-1 and Sentinel-2 images and DEM data by using Google Earth Engine. Then the river discharge is estimated by the rating curve. Meanwhile, the river discharge is evaluated using standard methods and compared with other products.

The increased availability and accuracy of recent remote sensing data accelerates the development of data products for hydrological modelling. Most hydrological models rely on the accurate representation of the Earth’s terrestrial surface including all waterways from small mountain streams to great lowland rivers in order to compute discharge. In light of this, the creation of the HydroSHEDS-X database, which is currently developed in an international collaborative project between the German Aerospace Center (DLR), McGill University, Confluvio Consulting, and the World Wildlife Fund, represents a new source for global digital hydrographic information. HydroSHEDS-X is the second version of the well-established HydroSHEDS database, which is freely available at https://hydrosheds.org. While the first version was derived from the digital elevation model (DEM) of the Shuttle Radar Topography Mission (SRTM), the foundation of HydroSHEDS-X are the elevation data of the TanDEM-X mission (TerraSAR-X add-on for Digital Elevation Measurement), which was created in partnership between the German Aerospace Center (DLR) and Airbus. HydroSHEDS-X benefits from the higher resolution of the underlying TanDEM-X DEM given its resolution of 0.4 arc-seconds worldwide including regions with latitudes higher than 60° North, which are not covered by the SRTM DEM. Details of this high-resolution DEM are preserved in the HydroSHEDS-X dataset by applying enhanced pre-processing techniques. This pre-processing of the elevation data comprises DEM infills for invalid and unreliable areas, an automatic coastline delineation with manual quality control, the generation of an open water mask, and the reduction that are caused by distortions of vegetation and settlements. The pre-processed DEM is further treated at a resolution of 3 arc-seconds to obtain a hydrologically conditioned DEM. Derived from this hydrologically conditioned version of the DEM, the HydroSHEDS-X core products comprise flow direction and flow accumulation maps as gridded datasets. The core products are complemented with secondary information on river networks, lake shorelines, catchment boundaries, and their hydro-environmental attributes in vector format. Finally, the database is completed with associated products. Available in standardized spatial units and at multiple scales starting from a resolution of 3 arc-seconds, HydroSHEDS-X is fully compatible with its original version and thus provides a consistent and easy-to-use database for hydrological applications from local to global scale. The main release of HydroSHEDS-X is scheduled for 2022 under a free license.

River plumes are well-visible on satellite imagery as sharp fronts (optical and thermal IR), resulting from the different properties of river vs ocean water bodies (e.g. temperature, salinity, sediment concentration). They serve as the link between river and ocean, transporting freshwater, (fine) sediments, nutrients and human waste (Halpern et al., 2008, Dagg et al., 2004; Joordens et al., 2001). Understanding the river plume dynamics will help to understand the transport of these substances. Moreover, it can provide valuable information on the river system. River plumes are formed by river freshwater entering the ocean, creating buoyant bodies of brackish water overlaying saltier sea water. The local buoyancy input of river freshwater interacts with tides, wind and waves. When the freshwater buoyancy is dominant, the system is stratified. This can be detected on sea surface temperature (SST) images during summer as a sharp front, as the top layer of freshwater heathens faster than the surrounding sea water (Pietrzak et al., 2011). Oppositely, when the mixing processes (tides, wind, waves) are dominant, the system will be well-mixed. This results in more gradual changes in temperature and salinity. For tide-dominated systems, freshwater pulses enter the ocean during ebb, resulting in multiple strong fronts on optical images. Wind and tides govern the propagation speed of these fronts, their thickness and direction in which they move (Rijnsburger et al., 2018). Consequently, fronts and their propagation can give information about the dynamics of the system and which processes are dominant. As water properties like salinity, temperature and sediment concentration determine the water colour, we can detect fronts on optical images. In this research, we study the potential relationship between fronts and river plume characteristics. First, we are developing the algorithms to detect fronts on satellite images. We hypothesize that the scale of the river plume can be related to the river discharge. We investigate this relationship using discharge data and the detected fronts as measure for the scale of the river plume. Next, we will investigate methods for coupling fronts detected on satellite images to fronts in a 3D numerical model of a river plume. We hypothesize that model performance can be improved by using the information retrieved from satellite images. Information-sparse river systems can profit in particular, where the satellite images can provide a valuable source of information to improve the modelling and understanding of the river plume dynamics. In this research we make a first step towards this goal, by investigating possibilities to retrieve (quantitative) information from satellite images and methods to couple the information to model output.

References
Dagg, M., R. Benner, S. Lohrenz, and D. Lawrence (2004), Transformation of dissolved
and particulate materials on continental shelves influenced by large rivers: plume
processes, Continental Shelf Research, 24, 833–858

Halpern, B. S., S.Walbridge, K. A. Selkoe, C. V. Kappel, F.Micheli, C. D’Agrosa, J. F. Bruno,
K. S. Casey, C. Ebert, H. E. Fox, R. Fujita, D. Heinemann, H. S. Lenihan, E.M. P.Madin,
M.T. Perry, E. R. Selig, M. Spalding, R. Steneck, and R. Watson (2008), A global map of
human impact on marine ecosystems, Science, 319, 948–953.

Joordens, J.C.A., A. J. Souza, and A. Visser (2001), The influence of tidal straining and
wind on suspended matter and phytoplankton distribution in the Rhine outflow region,
Continental Shelf Research, 21, 301–325

Pietrzak, J. D., de Boer, G. J., & Eleveld, M. A. (2011). Mechanisms controlling the intra-annual mesoscale variability of SST and SPM in the southern North Sea. Continental Shelf Research, 31(6), 594–610. https://doi.org/10.1016/j.csr.2010.12.014

Rijnsburger, S., Flores, R. P., Pietrzak, J. D., Horner-Devine, A. R., & Souza, A. J. (2018). The Influence of Tide and Wind on the Propagation of Fronts in a Shallow River Plume. Journal of Geophysical Research: Oceans, 123(8), 5426–5442. https://doi.org/10.1029/2017JC013422

Rijnsburger, S. (2021). On the dynamics of tidal plume fronts in the Rhine Region of Freshwater Influence. https://doi.org/10.4233/uuid:279260a6-b79e-4334-9040-e130e54b9360

Validation, including the determination of measurement uncertainties, is a key component of a satellite mission. Without adequate validation of the geophysical retrieval methods, processing algorithms and corrections, the computed geophysical parameters derived from satellite measurements cannot be used with confidence and the return on investment for the satellite mission is reduced. In this context, and in anticipation of the operational delivery of dedicated inland water products based on Copernicus Sentinel-3 measurements, the St3TART ESA project (Sentinel-3 Topography mission Assessment through Reference Techniques), is aimed at preparing a roadmap and providing a preliminary proof of concept for the operational provision of Fiducial Reference Measurements (FRM) in support of the validation activities of the Copernicus Sentinel-3 (S3) radar altimeter over land surfaces of interest (inland water bodies, sea ice and land ice).

In the framework of this project, the activities related to hydrology include a review of existing methodologies and associated ground instrumentation for validating and monitoring the performance and stability of the Sentinel-3 altimeter measurement via FRM. Methodologies and procedures are defined considering the errors and uncertainties coming from the point information of in-situ sensors, satellite measurements and the environment of the validation site. Based on these protocols and procedures, a roadmap is prepared in view of the operational provision of FRM to support the validation activities and foster exploitation of the Sentinel-3 SAR altimeter Land data products, over inland waters.
Then field campaign implementation and realization will be performed as a demonstrator based on the procedures and protocols defined and the roadmap.

In this ongoing project, a comprehensive review of altimeter uncertainties over inland water bodies was carried out on the basis of a literature review, leading to the identification of the different sources of error with their associated uncertainty level. A full review of all the sensors that have been used for many years for Cal/Val activities over inland waters has also been performed, combined with an analysis of innovative sensors that can fulfil the needs and potentially be used in the framework of the St3TART project. Cal/Val “super-sites” have been selected as demonstrators of the roadmap for operational FRM provision. We propose here to present the status and first results of these hydrology activities.

The Fully-Focused SAR (FF-SAR) processing, introduced in Egido and Smith (2016) allows obtaining a maximum resolution of 0.5 m in the along-track direction. It provides significant benefits for inland water altimetry investigations allowing the successful investigation of very small rivers and canals (Kleinherenbrink, 2020) that are typically harder to be analysed by using unfocused Delay-Doppler SAR (DD-SAR) data (about 300 m resolution in the along-track direction).
In its development, two major limitations were associated with the FF-SAR processing: 1) the presence of evenly spaced high sidelobes in the Point Target Response (PTR) due to the closed-loop burst mode implemented in Sentinel-3 & Cryosat-2 altimeter payloads, used for initial FF-SAR investigations, and 2) the heavy computational burden with respect to the unfocused DD-SAR processing.
The first limitation can be overcome by designing the radar system differently adopting an open-loop transmission scheme as, for instance, the one implemented in the altimeter payload of the Sentinel-6 Michael Freilich mission, launched on 21 November 2020 and operating since 21 June 2021.
The second limitation has been addressed in research works following Egido and Smith (2016) indicating that an improvement in terms of computational burden can be achieved by adopting algorithms in the frequency domain (Guccione et al., 2018).
Being the role of FF-SAR for future inland water altimetry well understood, along with the possibility to see it implemented with reduced drawbacks during the Sentinel-6 Michael Freilich mission, a collaboration has started between the ESA Altimetry Team, already hosting the successful SARvatore services portfolio for unfocused SAR & SARin altimetry, and Aresys.
Aresys has developed a generic FF-SAR prototype processor, that is able to process data acquisition from different instruments and exploiting the frequency-domain Omega-K algorithm (Guccione et al., 2018 & Scagliola et al., 2018). The Aresys's FFSAR prototype processor for CryoSat-2 allows users to process, on line and on demand, low-level CryoSat FBR products in SAR mode up to FF-SAR Level-1 products with self-customized options. Additionally a wide set of processing parameters is configurable, allowing as an example to select the along-track resolution or to obtain FFSAR multilooked waveforms at the desired posting rate.
The collaboration led to the creation of a new service for the processing of CryoSat-2 data in FF-SAR mode. Users will be able to select the following options: 1) range oversampling factor, 2) bandwidth factor (responsible for the along-track resolution value) and 3) multilook posting rate (1Hz-500Hz). Geophysical corrections and L2 estimates from both a threshold peak retracker and an ALES-like subwaveform retracker are part of the output package. In preliminary open ocean analyses, very good results on SSH noise have been obtained by the ALES-like subwaveform retracker.
In this presentation, the Aresys FF-SAR prototype processor is described and the outcome of some preliminary validation activities, performed by a group of altimetry researchers, is reported.
The service, to be soon extended to allow the processing of Sentinel-3 and Sentinel-6 data, will be made available to the altimetry community in early 2022 as part of the Altimetry Virtual Lab, a community space for simplified services access and knowledge-sharing. It will be hosted on EarthConsole (https://earthconsole.eu), a powerful EO data processing platform now also on the ESA Network of Resources (info at altimetry.info@esa.int).

References
Egido A., Smith W. H. F., “Fully Focused SAR Altimetry: Theory and Applications” IEEE Transactions on Geoscience and Remote Sensing, Volume: 55, Issue: 1 , Jan. 2017, doi: 10.1109/TGRS.2016.2607122.
Kleinherenbrink M., Marc Naeije, Cornelis Slobbe, Alejandro Egido, Walter Smith, The performance of CryoSat-2 fully-focussed SAR for inland water-level estimation, Remote Sensing of Environment, Volume 237, 2020, 111589, ISSN 0034-4257, https://doi.org/10.1016/j.rse.2019.111589.
Guccione P., Scagliola M., and Giudici D., “2D Frequency Domain Fully FocusedSAR Processing for High PRF Radar Altimeters” Remote Sens. 2018, 10, 1943; doi:10.3390/rs10121943.
Scagliola M., Guccione P., “A trade-off analysis of Fully Focused SAR processing algorithms for high PRF altimeters”, 2018 Ocean Surface Topography Science Team (OSTST) Meeting. https://meetings.aviso.altimetry.fr/programs/complete-program.html.

In times of ever decreasing amount of in-situ data for hydrology, satellite altimetry has become key to provide global and continuous datasets of water surface height. Indeed, studying lakes, reservoirs and rivers water level at global scale is of prime importance for the hydrology community to assess the Earth’s global resources of fresh water.
Much progress has been made in the altimeters’ capability to acquire quality measurements over inland waters. In particular, the Open-Loop Tracking Command (OLTC) now represents an essential feature of the tracking function. This tracking mode’s efficiency has been proven on past missions and it is now stated as operational mode for current Sentinel-3 and Sentinel-6 missions. It has benefited from iterative improvements brought to onboard tables contents repeatedly since 2017.

In 2022, new updates will be performed on the onboard OLTC tables of the Sentinel-3A and Sentinel-3B missions, as well as Sentinel-6A and Jason-3 following their successful Tandem Phase.
The number of hydrological targets used to define the tracking command currently reaches an unprecedented number of targets of almost 100,000 for each Sentinel-3 and about 30,000 for Sentinel-6A. We expect to define a similar number of targets in the interleaved orbit of Jason-3, previously flown by Jason-2, although mostly in Closed-Loop Mode.
These major improvements over the last few years have been made possible by the analysis and merging of the most up-to-date digital elevation models (SRTM, MERIT and ALOS/PalSAR) and water bodies databases (HydroLakes, GRaND v1.3, SWBD, GSW, SWORD). In addition, special effort is put into introducing the most recent reservoir databases. This methodology ensures coherency and consistent standards between all nadir altimetry missions and types of hydrological targets.
Finally, additional efforts have been carried out to define a relevant tracking command outside of hydrological areas, in order to keep track of the continental surface and enabling potential other land applications, while optimizing the OLTC onboard memory.

The OLTC function of nadir altimeters constitutes a great asset for building a valuable and continuous record of the water surface height of worldwide lakes, rivers, reservoirs, wetlands and even a few continental glaciers.
This work is essential at institutional and scientific levels, to make the most of current altimeters coverage over land and to prepare for the upcoming calibration and validation of the Surface Water and Ocean Topography (SWOT) mission. In this context, we will present an overview of OLTC achievements and perspectives for future altimetry missions.

In a global warming and climate change context, populations all over the world are impacted by an increasing number of hydrological crisis (flood events, droughts, ...), mainly related to the lack of knowledge and monitoring of the surrounding water bodies. In Europe, flood risk accounts for 46% of the extreme hazards recorded over the last 5 years and current events confirm these figures for France and Europe. Although the main rivers are properly monitored, a wide set of small rivers contributing to flood events are not monitored at all. There is a clear lack of river basins monitoring with regard to the rapid increase of extreme events. In France, 20000 km of regulatory rivers are monitored in real time while 120000 km are required. Moreover, hydrological surveys are currently insured by heterogeneous means from a country to another and even inside a country, from a region to another. It results in a high cost level to deploy a robust, relevant and efficient monitoring of all watercourses at risks. Therefore there is a real need for affordable, flexible and innovative solutions for measuring and monitoring hydrological areas in order to address climate change and flood risk within the water big cycle.

vorteX.io offers an innovative and intelligent service for monitoring hydrological surfaces, using easy-to-install and fixed remote sensing in-situ instruments, based on compact light-weight altimeter inspired from satellite technology: the micro-stations. It provides in real time, with a high accuracy, hydro-meteorological parameters (water surface height, water surface velocity, images & videos) of the observed watercourses. The combination of these in-situ data with satellite measurements is thus optimal for downstream services related to water resources management and assessment of flood/drought risks. Thanks to the development of the innovative micro-station and the onboard processing using artificial intelligence algorithms, the vorteX.io solution will provide an anytime/anywhere real time hydro-meteorological database to prevent communities from flood risks and secure goods anywhere at any time. The solution thus aims to cover the whole Europe through a non-binding turnkey service to ensure the resilience of territories to climate change and guarantee the safety of people and goods.

It worth to mention that the vortex.io solution can also address the needs of in-situ measurements (Fiducial Reference Measurements) for Cal/Val activities on inland water bodies. Indeed, the vorteX.io micro-station is able to automatically wake up and perform measurements at the exact moment of the satellite overflight thanks to satellite ephemerides. With this feature, there is no time delay between in-situ measurements and the satellite overflight. Water heights are provided with respect to the ellipsoid or local geoid. All geophysical corrections required can be applied on the fly. Different hydrological variables are measured (water surface height, the associated uncertainty, water surface speed) and new others are planned to be added in the next future (water surface temperature, turbidity, …). The vorteX.io solution has already been used in various CNES and ESA projects and will be implemented in the ESA St3TART project and will be used for Cal/Val activities of the future SWOT mission on the Garonne river.

Water level time series based on satellite altimetry over rivers is to a large degree limited to solutions at virtual stations, the locations where the ground tracks repeatedly intersect the river. Such a paradigm has prevented the community from exploiting satellites in geodetic orbit like CryoSat-2, SARAL/AltiKa to their full potential. Additionally, we are in a unique situation with an unprecedented number of missions that together give a much more detailed picture in space and time of the water level, than can be achieved at virtual stations.

An alternative to the virtual stations approach is the so-called reach-based method, where the water level is reconstructed based on available data within a river reach. A reach-based approach has the advantage that the water levels are seen in a context, as a river elevation profile can be formed. This makes it possible to detect blunders, which is more challenging at virtual stations without prior knowledge, where only a few observations may be available. Additionally, a reach-based approach is not affected by tracks that intersect the river at a small/large angle, which typically will degrade the result at a virtual station.

However, combing noisy water levels at different locations, times, and acquired from different missions is indeed challenging. Here we present a new reach-based method to reconstruct the river water level time series. We model the observations given in 1D space and time as a Gaussian Markov Random Field. In the model, we account for inter-mission bias, satellite-dependent noise, and use and increasing spline function to represent a time-independent water level along the reach.

Here, we demonstrate the new method for different river reaches from, e.g., the Missouri and Mississippi Rivers, and validate the result against in situ data. We show that the model is able to reconstruct the water level for reaches with different hydrological regimes, e.g., the presence of reservoirs.

The research for this work was partly funded by the ESA Permanent Open Call STREAMRIDE and RIDESAT Projects.

Surface water level and river discharge are key observables of the water cycle and among the most sensible indicators that integrate long-term change within a river basin. As climate change accelerates and intensifies the water cycle, streamflow monitoring allows the understanding of a broad range of science questions focused on hydrology, hydraulics, biogeochemistry, water resources management and flood protection. Streamflow change is a response to anthropogenic, as deforestation, land use change, urbanization, and natural, as climate modes, climate variability, rainfall, processes. Climate and internal drainage mechanisms affect and control, together with river discharge also lake, reservoirs, mountain glacier storage. An enhanced global warming, predicted by coupled models as due to anthropogenically-induced greenhouse warming, is expected to accelerate the current glacier decline. Moreover, precipitation cause fluvial floods, when rivers burst their banks as a result of sustained or intense rainfall, and pluvial floods, when heavy precipitation saturates drainage systems.
Change in storage and release of water is important for watershed management, including the operation of hydroelectric facilities and flood forecasting, and have direct economic effects. We analyse the observability of extremes water levels events (in low/high water and in discharge) and the long-term variability from space data for the Rhine, Elbe River catchments in central Europe. We consider for the river Rhine the extreme in July 2021 in particular.

Over the last decade, the merging of innovative space observations with in-situ data provides a denser and accurate two-dimensional observational field in space and time compared to the previous two decades, and allow to better monitor the impact of water use and characterize climate change. The new generation of space borne altimeters includes Delay Doppler, laser and bistatic SAR altimeter techniques. The central hypothesis is that these new observations outperform conventional altimetry (CA) and in-situ measurements providing (a) surface water levels and discharge of higher accuracy and resolution (both spatial and temporal), (b) new additional parameters (river slope and width) and (c) better sampling for flood event detection and long-term evolution, providing valuable new information to modelling.

In this study, radar and laser satellite altimetry and satellite images provide the space observations, radar altimeter data are processed by the GPOD ESA service. Time-series of Water Surface Elevation (WSE) are built by two methods. In the first method, time-series are built by collecting the observations of one single virtual point (one-VS), while in the second time-series are constructed from observations at multiple VS (multi-VS) after correcting for the river mean slope. The accuracy of the time-series built with both methods is 10 cm and 30 cm for Sentinel-3. The second method applied to CryoSat-2 SARIN data produces less accurate time-series and gives a similar accuracy for unfocused and fully focused (FF-SAR) processing. The impact on the results of the chosen centerline and mean river slope is investigated using the SWORD database and the national agency BfG database. The river discharge is evaluated.

In the long run, the long-term variability of the combined altimetric and in-situ water level and river discharge time-series depends on the changing climate and correlate with temperature and precipitation at basin and regional scales.

This study is part of Collaborative Research Centre funded from the German Research Foundation (DFG): “Regional Climate Change – The Role of Land Use and Water Management”, in sub-project DETECT-B01 “Impact analysis of surface water level and discharge from the new generation altimetry observations” which addresses the two research questions: (1) How can we fully exploit the new missions to derive water level, discharge, and hydrodynamic river processes, and (2) can we separate natural variability from human water use?

Water resource management is critical in many arid environments. The understanding and modelling of hydrological systems shed light on important factors affecting scarce water resources. In this study, a semi distributed hydrological model capable of simulating water balance in large geographical catchments and sub basin was used for runoff estimation in the Okavango Omatako catchment in Namibia. The model was configured for a thirty–one year period from 1985 to 2015 as per the availability of data for the study area. Subsequently, calibration and validation processes followed for the period 1990-2003 (calibration) and 2004-2008 (validation) using the sequential uncertainty fitting 2 (SUFI 2) algorithm. For the evaluation of the simulation of the Okavango Omatako catchment, two methods were used: i. model prediction uncertainty and ii. model performance indicators. Prediction uncertainty was used to quantify the goodness of fit between observed and simulated result of model calibration, which is measured by P factor and R factor. The P-factor achieved 0.77 during calibration and 0.68 for the validation. The value for calibration was adequate while the validation value was around the recommended value of 0.7. On the other hand, the R factor attained 1.31 in the calibration and 1.82 during validation. The calibration result was within the acceptable range while the validation was slightly on the upper side. The following indicators were used to evaluate the model performance through calibration and validation results respectively; Nash Sutcliffe Efficiency (NSE) with 0.82 and 0.80, Coefficient of determination (R^2) with 0.84 and 0.89, Percent bias (PBIAS) achieving -20≤PBIAS≤-1.1 and residual variation (RSR) performing 0.42 and 0.44. All performance indices achieved very good ratings apart from PBIAS validation which rated as satisfactory. It is therefore recommended to use SWAT for semi-arid streamflow simulations as it demonstrated reasonable results in modeling high and low flows.

Using satellite altimetry over poorly gauged basins where in situ data are scarce can be very beneficial for river monitoring, which is becoming more important due to increasing challenges with managing freshwater resources in a world affected by climate change and economic growth. As the resolution of satellite altimeters increases, the potential for their use grows. When CryoSat-2 was launched by ESA in 2010, the 300 m along track resolution of the Synthetic Aperture Radar (SAR) data allowed for the study of rivers much narrower compared to what was possible for missions such as Envisat, where only Low Resolution Mode (LRM) data were available. However, the resolution of SAR altimetry is still not high enough to monitor narrow rivers and rivers in mountainous areas.
In recent years, the Fully Focused SAR (FF-SAR) processing has been used to increase the along-track resolution further, all the way down to half the antenna length (Egido et al., 2017). The FF-SAR processing can be applied to all SAR altimeter missions, i.e. CryoSat-2, Sentinel-3 and Sentinel-6/Jason-CS. It has previously been shown that that FF-SAR processing can be used to obtain water levels for objects of just a few meters in width (Kleinherenbrink et al., 2020).

Satellite altimetry also includes height measurements from lidar measurements. In 2018, NASA launched the ICESat-2 satellite carrying the Advanced Topographic Laser Altimeter System (ATLAS) which uses a green laser to estimate the distance between the satellite and the point of reflection on the ground. ATLAS detects every single photon that finds its way back to the instrument after reflection. The along-track resolution of ICESat-2 is around 0.7 m but depends on the number of detected photons. For highly specular surfaces the resolution is much higher, and in some cases it might be lower.

Here, we compare the respective pros and cons of FF SAR Sentinel-3 and ICESat-2 altimetry over the Yellow River basin in China and other rivers that are challenging for SAR and LRM altimetry.

We present river levels derived from Sentinel-3 data using the processor provided by the SMAP FFSAR CLS/ESA/CNES project and river levels from the ATL03 and ATL13 ICESat-2 products and compare these with available in situ data.

Operational hydrologists in Czechia often need information on the position of the zero isochion (otherwise known as snow line) in order to correctly delineate the snow-covered area in geomorphological regions. This activity is extremely helpful when determining the information on water stored in snow during the winter season, which, in turn, helps the Czech hydrologists properly model the expected runoff or to better quantify the individual components of water balance. So far, the estimation of the spatial distribution of snow cover and snow water equivalent has been performed through spatial interpolation which is supposed not to estimate the positive values below the zero isochion, the altitude of which is calculated once a week using the combination of in-situ data collected by the Czech Hydrometeorological Institute and remotely sensed data coming from the MODIS imagery. The advantage of the MODIS products is the time resolution, while their disadvantage is the resolution in the space domain (i.e. pixel of the edge of 500 m). This disadvantage often plays a role in discriminating between various classes of the landscape and often prevents the recognition of the snow-covered areas in forests. Therefore, the purpose of this contribution is to experimentally employ another satellite product here, which has a finer spatial resolution, that is the COPERNICUS Sentinel-2 data. The R package 'sen2r' was used to download the Sentinel-2 images and also to further process the data to get the Normalized Difference Snow Index, based on which the discrimination between the snow-covered areas and the areas without snow was carried out for the territory of Czechia (or at least its selected regions) and for the winter season defined by the months of November through May. The task is to find out if the Sentinel-2 data can be routinely used instead of the MODIS data when defining the position of the zero isochion in Czechia. The by-product of the analyses might be the substitution of the commercial processing software by the selected open source software.

While multimission satellite altimetry over inland waters has been known and used for more than two decades, monitoring of lakes and reservoirs is far from fully operational. Depending on the spatiotemporal coverage offered by altimetry missions, the concept has its fundamental limitations. However, for medium to large water bodies where altimetry can provide meaningful information, the multimission is still hampered by what is known as inter-satellite bias. Studies have been performed to quantify absolute altimetry biases at calibration sites and relative altimetry biases on a global scale. However, a thorough understanding of the biases between satellites over inland waters has not yet been achieved.

We explore the possibility of resolving the biases between satellites over lakes and reservoirs. Our solution for estimating the biases between overlapping and non-overlapping time series of water levels from different missions and tracks is to rely on the time series of surface area derived from the satellite imagery. The area estimated by the imagery acts as an anchor for the water level variations, making the area-height relationship the basis for estimating the relative biases. We estimate the relative biases by modeling the area-height relationship within a Gauss-Helmert model conditioned on an inequality constraint. For the estimation, we use the expectation maximization algorithm that provides a robust estimate by iteratively adjusting the weights of the observations.

We evaluate our method on a limited number of lakes and reservoirs and validate the results against in situ water level data. Our results show the presence of inter-satellite and also inter-track biases at the decimeter level, which are different from the global bias estimates.

With the increasing demography, inland water is a more and more pressured resource for the population needs as well as a societal risk for local populations. It is also a fundamental element for industry and agriculture, therefore becoming an economic and political stake. The monitoring of inland water level, proxy to freshwater stocks, conditions of navigability on inland waterways, discharge, flood prevention, is thus an important challenge. With the decreasing number of publicly available in situ water level records, the altimetry constellation brings a powerful and complementary alternative.

The Copernicus services provide operational water level timeseries products based on satellite altimetry data and their associated uncertainties over inland waters worldwide. The Copernicus Global Land service delivers near real time timeseries updated daily over both river and lakes, the Copernicus Climate Change service (C3S) focuses on lakes, with data updated twice a year. The number of operational products is in constant augmentation since 2017 thanks to the combined effort of CNES (THEIA Hydroweb and SWOT Aval projects) and Copernicus projects.

Evolutions to successively integrate new missions are regularly performed: the Sentinel-3A and 3B missions allowed to define new targets as well as exploiting the successive upgrades of onboard Open Loop Tracking Commands that ensure the altimeters hooking on the water targets have been performed. This yielded an operational monitoring of more than 11000 virtual stations over rivers and more than 180 lakes worldwide (as of 2021). The services will also integrate Sentinel-6A in 2022 to ensure the continuity of the timeseries which is essential to ensure continuity of the long lakes timeseries under the Topex/Jason ground track. This is of particular importance for the C3S long lake water level timeseries.

This presentation will detail both the processes yielding the definition of new targets and their qualification for operation as well as the regular quality assessment of the produced water level timeseries. The metrics and associated results will be detailed based both on intra satellite comparisons and on Insitu datasets. In particular, the benefits of recent evolutions of the services will be stressed: data precision improvement brought by the SAR mode used onboard S3A&B and continuity over the long term Topex/Jason timeseries, with the Sentinel-6A mission, which is essential for climate purposes. A first insight will be given on the future further improvements of the services products, expected with the ingestion of the new Inland Water products from the Thematic Instrument Processing Facilities (T-IPF) currently under development in ESA Mission Performance Cluster.

Estimates of the spatio-temporal variations of Earth’s gravity field based on the Gravity Recovery and Climate Experiment (GRACE) mission observations have shed a new light into large scale water redistribution at inter-annual, seasonal and sub-seasonal timescales. As an example, it has been shown that for many large drainage basins the empirical relationship between aggregated Terrestrial Water Storage (TWS) and discharge at the outlet reveals an underlying dynamics that is approximately linear and time-invariant (see attached figure for the Amazon basin).
We built on this observation to first put forward lumped-parameter models of the TWS-discharge dynamics using a continuous-time linear state-space representation. The suggested models are calibrated against TWS anomaly derived from GRACE data and discharge records using the prediction-error-method. It is noteworthy that one of the estimated parameters can be interpreted as the total amount of drainable water stored across the basin, a quantity that cannot be observed by GRACE alone. Combined with the equation of water mass balance, these models form a consistent linear representation of the basin-scale rainfall-runoff dynamics. In particular, they allow to derive analytically a basin-scale instantaneous unit hydrograph. We illustrate and discuss in more detail the results in the case of the Amazon basin and sub-basins, which present relatively simple TWS-discharge dynamics well approximated by first-order ordinary differential equations. Finally, we briefly discuss how to refine the linear models by introducing non-linear terms to better capture delays and saturations.
With such linear and non-linear models at hands, it is possible to use classical Bayesian algorithms to filter, smooth or reconstruct the basin aggregated TWS and/or discharge in a consistent manner. As such, we claim that these lumped models can be an alternative to more complex and spatially distributed hydrological models in particular for TWS and discharge time series reconstruction. We also briefly examine the conditions under which the linear models can be used to do hydrology backwards, that is, estimating simultaneously the TWS and the unknown input precipitation minus evapotranspiration from discharge records.

Sentinel-3 is an Earth observation satellite series developed by the European Space Agency as part of the Copernicus Programme. It currently consists of 2 satellites: Sentinel-3A and Sentinel-3B, launched respectively on 16 February 2016 and 25 April 2018. Among the on-board instruments, the satellites carry a radar altimeter to provide operational topography measurements of the Earth’s surface. Over Inland waters, the main objective of the Sentinel-3 constellation is to provide accurate measurements of the water surface height, to support the monitoring of freshwater stocks. Compared to previous missions embarking conventional pulse limited altimeters, Sentinel-3 is measuring the surface topography with an enhanced spatial resolution, thanks to the on-board SAR Radar ALtimeter (SRAL), exploiting the delay-Doppler capabilities.
To further improve the performances of the Sentinel-3 Altimetry LAND products, ESA is developing dedicated and specialized delay-Doppler and Level-2 processing chains over (1) Inland Waters, (2) Sea-Ice, and (3) Land Ice areas. These so-called Thematic Instrument Processing Facilities (T-IPF) are currently under development, with an intended deployment by mid-2022. Over inland waters the T-IPF will including new algorithms, in particular the hamming window and the zero-padding processing. Thanks to the hamming window, the waveforms measured over specular surfaces are cleaned from spurious energy spread by the azimuth impulse response. The zero-padding provides a better sampling of the radar waveforms, notably valuable in case of specular energy returns.
To ensure the missions requirements are met, ESA has set up the S3 Land Mission Performance Cluster (MPC), a consortium in charge of the qualification and the monitoring of the instrument, and core products performances. In this poster, the Expert Support Laboratories (ESL) of the MPC present a first performance assessment of the T-IPF level-2 products over inland waters. The analyses presented are made over a large set of worldwide lakes and rivers. Comparisons with InSitu datasets, for example benefiting from the contribution of the St3TART project, will provide an estimate of the topography precision, and will be discussed for rivers and lakes of various sizes. Inter satellite comparisons are also in the scope of the studies and Water Surface Height estimates consistency in between Sentinel-3 and ICESat-2 will complement this analysis.
The quality step-up provided by the hydrology thematic products, and highlighted in this poster, is a major milestone. Once the dedicated processing chain is in place for the inland waters acquisitions, the Sentinel-3 STM level-2 products will evolve and improve more efficiently over time to continuously satisfy new requirements from the Copernicus Services and the scientific community.

In the frame of the ESA HYDROCOASTAL project, led by Satoc Ltd, a Test Dataset (TDS) is being developed by partners starting from Altimetry L1A data products (Sentinel-3, CryoSat-2) up to L2 covering the coastal and inland water domains. Then, specific products are targeted to the monitoring of inland water: L3 for the river and lake water level estimations and L4 for the estimation of river discharge.

The TDS is a test-benchmark run focusing on selected region of interest. It serves to perform extensive validation activities in the objective to qualify and quantify the quality of the various output products (L2, L3 and L4). Outcomes of this activity will serve to validate the methods and algorithms to be adopted by the team before the project initiates the production of Globally Validated Products (GVP).

This presentation focuses on the results obtained at L3 for the estimation of river water level.

Partners involved in the production of the L2 products all have developed and implemented specific waveform retracking algorithms. Outputs from each of these retrackers, limited to those implemented over the inland water domain, are processed further on to produce L3 river water level estimates.

Results will describe the performance of each of the L2 retrackers, from the L3 "point of view". The analysis will not only focus on the vertical accuracy of the data but also on the ability of the complete L1-to-L3 chain to produce consistent water level time series, considering the effective temporal sampling as another key indicator of the data quality.

The validation activity is done over the Amazon basin and involves the systematic comparison to in situ data from the ANA (Agência Nacional de Águas, Brazil).

Eventually, the results will highlight the strength and possible weaknesses of the retracking algorithms, helping to decide which retracker is eligible to be implemented in the production of the ultimate GVP datasets.

Across Iran, extraction of non-renewable groundwater has sparked water-related stress, increased salinisation of groundwater sources, and accelerated ground subsidence (Olen, 2021).

Both local and regional scale land-surface deformation has resulted from the decline in groundwater levels (Motagh et al., 2008). Moreover, the gap between groundwater use and renewal is so large that the resulting short-term impacts are likely to be irreversible (Olen, 2021). Quantifying the extents and rates of deformation related to groundwater extraction could therefore inform groundwater management approaches.

Here we present a catalogue of around sixty major, currently subsiding basins within the political borders of Iran. We use the COMET LiCSAR automated processing system to process seven years (2015-2021) of Sentinel-1 SAR acquisitions. The system generates short baseline networks of interferograms. We also correct for atmospheric noise using the GACOS system (Yu et al., 2018) and perform time-series analysis using open-source LiCSBAS software (Morishita et al., 2019) to estimate the cumulative deformation.

We also present vertical and horizontal velocity components of basin subsidence obtained through the decomposition of line-of-sight InSAR velocities. Subsiding basins are characterised and catalogued using the resulting interferogram time-series based upon the extents and rates of vertical motion associated with basin and agricultural areas. LiCSBAS time series analysis reveals maximal vertical rates of subsidence reach thirty-six centimetres per year in basins north-west of Tehran.

Finally, we present and demonstrate a beta version of the COMET-LiCS Sentinel-1 InSAR Subsiding Basin Portal. The portal aims to provide tools for the online analysis of automatically processed LiCSAR Sentinel-1 interferograms and subsequent LiCSBAS timeseries. The portal’s tools are designed to allow key stakeholders to search quickly through processed imagery and make critical assessments related to the extents and rates of basin subsidence. Initially the portal characterises Iranian basins but will increasingly have a global focus. Ongoing updates will be made to the portal’s interferograms and timeseries extended as more Sentinel-1 data is acquired.

Future work will focus on determining which basins are experiencing accelerating or decelerating subsidence rates. Ultimately, our quantification of ground deformation- particularly subsidence-related to groundwater withdrawal- could contribute to the development of a wider framework for monitoring complex risk pathways in similar, water stressed regions.

Subsidence is defined as gentle and graduate land surface lowering or collapse, which can be caused by either natural or anthropogenic activities. Land subsidence slowly proceeds due to sediment compaction under the pressure of overlying sediments. More intense motions in amplitude and time can be induced or worsened by human activity, such as groundwater withdrawal or underground mining.
Conventional geodetic measurement techniques have long been used for monitoring deformation processes. In particular, several methods including levelling, total station surveys, and GPS are still currently used for subsidence detection and monitoring. Nevertheless, these techniques are suitable only for local and site-specific analysis as they measure subsidence on a point-by-point basis, requiring a dense network of ground survey markers.
Space-borne radar interferometry approach allows measuring ground movement between two radar images acquired at different times on the same area, on a pixel-by-pixel basis. It is therefore remotely sensed, covering wide areas, quicker, and less labour intensive compared with conventional ground-based survey methods. More recently, radar interferometry rapidly growth and became a well-established Earth observation technique. The last decades witnessed a large exploitation of satellite InSAR (Interferometric Synthetic Aperture Radar) data.
The completeness of Web of Science (WoS) database was exploited to collect and critically review the current scientific state-of-the-art in the field of subsidence analysis using satellite InSAR in the last thirty years. From the pioneering work dating back to late nineties, the use of InSAR dramatically increased and thanks to the technological advances in terms of both acquisition sensors and processing algorithms, it is now possible to cover all the analysis of subsidence-related deformation at different stages, such as detection and mapping, monitoring and characterization, modelling and simulation.
This work is aimed to illustrate the role of satellite interferometry for subsidence analysis at a worldwide level over the last 20 years, highlighting current applications and future perspectives. From the WoS database original articles, book chapters, conference proceedings, and extended abstracts written in English and published by international journals after peer review where the authors exploited the InSAR technique to study subsidence were gathered. The data collection involved all contributions referring to every area in the world, looking for state by state. The data collection was operated in May 2021 and a final list of 766 contributions was retained. After the first skimming, each article was read and critically analysed in order to extract several information and identify the irrelevant contributions automatically extracted by the WoS database by the advanced search parameters. After this depth analysis, 73 contributions were further deleted by the list because they are related to volcanic systems, earthquakes or faults moving, or dam structures. In the end, the database was filled by the information extracted by 693 contributions.
Beside the general information about the selected contributions automatically downloaded from WoS (e.g., Publication Type, Authors and related affiliations, article Title, etc), it was necessary to insert new fields to catalogue the article to obtain a more detailed characterization:
- “CU” - Country Region of the authors;
- Area of Interest - localization of the study area investigated in the analysed work;
- SAR Satellite - list of the SAR satellite used to develop the investigation;
- Cause - list of the triggering factor of the subsidence as indicated in each contribution;
- Processing Technique - the processing technique adopted to retrieve ground deformation;
- Applications - the presented work were categorized according to the aim of the study
- Integration - the validation or integration, if any, with other data;
- Field evidences - the recording of damage on structure or ground.
The literature review highlighted that subsidence analysis covers 62 countries with at least one case study, by corresponding authors coming from 46 different nations. All the continents are covered, with the exception of Antarctica. The most represented country is China with 258 applications, followed by USA, Italy e Mexico with 59, 53 and 43 applications, respectively (Figure 1).
All the radar imagery archive were exploited to collect past and recent information on ground subsidence. The most used satellite platform turned out to be the C-band Envisat (counting 286 applications), followed by Sentinel-1 (154), ERS (153), L-band ALOS-1 (126), and X-band TerraSAR-X (95).
Concerning the image processing techniques adopted, either DInSAR (Differential InSAR) and InSAR algorithms were exploited, with 162 and 590 applications, respectively. Among the InSAR approaches SBAS (Small BAseline Subset) is represented by 187 case studies, PSInSAR (Persistent Scatterer Interferometry) is used in 146 circumstances while IPTA (Interferometric Point Target Analysis) is used within 36 contributions.
The triggering factors which resulted in land subsidence can be distinguished in two main groups, anthropogenic and natural. The first category counted 716 contributions, distributed mainly in groundwater exploitation (383), mining activities (137), urban loading (105), while in the second category the most represented factors are sediment compaction (117) and tectonic deformation (19).
Finally, the main aim of each contribution was identified and as a result 290 works dedicated to the monitoring of the subsidence phenomena, 171 devoted to the precise mapping of the extent of the vertical displacement and in 78 cases InSAR data were implemented for the modelling of the deformation were counted.
Therefore, satellite InSAR largely demonstrated to be highly valuable for subsidence studies in different setting and for different purposes, providing insights on slow-moving subsidence deformation mechanism. The upcoming European–wide EGMS (European Ground Motion Service), whose first baseline release is foreseen for the end of 2021, will represent a fundamental support for land subsidence analysis, providing a valuable flood of information regarding surface vertical deformation.

Ground motion, such as land subsidence, can be due to human (groundwater extraction, artificial loading) and natural causes. The latter are related to the geological setting, properties of soil and also to climatic stress (drought periods) and they can affect human-induced subsidence. It is possible that more than one cause contributes to the ground deformation, and it can be difficult to determine and quantify the contribution of each cause. In addition, also the socio-economic development factors, such as the increase in water demand and urbanization and population growth, can contribute to worsening subsidence, especially regarding subsidence due to groundwater extraction which is often overexploited. InSAR data can be a valid support in the study of the ground movements, providing valid products (starting from time series up to displacement maps) that can cover wide areas even where in situ monitoring instruments may be missing. This work will focus on the analysis of A-DInSAR time series by applying several methodologies and additional factors, such as analysis of topography, lithology, land use and geological setting, will be taken into account. In particular, ONtheMOVE methodology (InterpolatiON of InSAR Time series for the dEtection of ground deforMatiOn eVEnts) will allow to classify the A-DInSAR TS trend (uncorrelated, linear, non-linear) and to identify areas with non-linear target clusters. Wavelet analysis and Independent Component Analysis will be performed both on A-DInSAR data and piezometric data in order to unravel and correlate the main components of both TS. Satellite data that will be used cover a period from 2015 to 2021 in a test site in Brescia province (Lombardia region, Italy), in which subsidence is related not only to groundwater extraction but also to compaction of clay, peat oxidation and compaction due to artificial loading. The results of this study will contribute to improve the knowledge of ground deformations in the test site, and they will be helpful in the characterization of aquifer parameters to fill gaps in data especially when in situ monitoring systems are scarce.

The Willcox Basin, located in southeast of Arizona, USA, covers an area of approximately 4,950 km2 and is essentially a closed broad alluvial valley basin. The basin measures approximately 15 km to 45 km in width and is 160 km long. Long-term excessive groundwater exploitation for agricultural, domestic and stock applications has resulted in substantial ground subsidence in the Willcox Groundwater Basin. The land subsidence rate of the Willcox Basin has not declined but has rather increased in recent years, posing a threat to infrastructure, aquifer systems, and ecological environments.
In this study, spatiotemporal characteristics of land subsidence in the Willcox Groundwater Basin was first investigated using an interferometric synthetic aperture radar (InSAR) time series analytical approach with L-band ALOS and C-band Sentinel-1 SAR data acquired from 2006 to 2020. The overall deformation patterns are characterized by two major zones of subsidence, with the mean subsidence rate increasing with time from 2006 to 2020. The trend of the InSAR time series is in accordance with that of the groundwater level, producing a strong positive correlation (≥0.93) between the subsidence and groundwater level drawdown, which suggests that subsidence is a result of human-induced compaction of sediments due to massive pumping in the deep aquifer system and groundwater depletion.
In addition, the relationship between the observed land subsidence variations and the hydraulic head changes in a confined aquifer in accordance with the principle of effective stress and hydromechanical consolidation theory. Therefore, integrating the InSAR deformation and groundwater level data, the response of the aquifer skeletal system to the change in hydraulic head was quantified, and the hydromechanical properties of the aquifer system were characterized. The estimated storage coefficients, ranging from 〖6.0×10〗^(-4) to 0.02 during 2006-2011 and from 〖2.3×10〗^(-5) to 0.087 during 2015-2020, signify an irreversible and unrecoverable deformation of the aquifer system in the Willcox Basin. The reduced average storage coefficient (from 0.008 to 0.005) indicates that long-term overdraft has already degraded the storage ability of the aquifer system and that groundwater pumping activities are unsustainable in the Willcox Basin. Historical spatiotemporal storage loss from 1990 to 2020 was also estimated using InSAR measurements, hydraulic head and estimated skeletal storativity. The estimated cumulative groundwater storage depletion was 〖3.7×10〗^8 m3 from 1990 to 2006.
Understanding the characteristics of land surface deformation and quantifying the response of aquifer systems in the Willcox Basin and other groundwater basins elsewhere are important in managing groundwater exploitation to sustain the mechanical health and integrity of aquifer systems.

Groundwater has been extracted in the municipality of Delft since 1916. The extraction used to be owned by a privately owned yeast factory, but when the company recently seized production, the extraction was transferred to the municipality of Delft. Though the extraction has no functional use at this time, the city of Delft is worried that due to the size of the extraction, which is currently 1200 m3 (about an olympic swimming pool) per hour, stopping the extraction will have a big effect on the buildings in the historic 16th century city centre. The current annual costs of water disposal for the municipality are €2.5 million.

Since 2017 the municipality is slowly phasing out the extraction. The reduction in groundwater extraction must be carefully controlled to avoid an abrupt rise of the surface level due to swelling of the ground and consequent damage to infrastructure. To monitor the effects of the reduction, extensive measurements like groundwater levels have been collected since 2010.

It is estimated that the shutdown of these wells over a 30 year period could lead to ground swelling of more than 10 cm near the extraction wells. Abrupt and uneven swelling can cause damage to buildings, and infrastructure like tunnels, parking garages etc. There are 70,000 buildings in the 5 km surrounding the extraction site, including a range of irreplaceable historic buildings.

To supplement the groundwater level measurements, InSAR is used to monitor the changes in the uplift of the soil (since 2014) and guide the speed with which the groundwater extraction is reduced. Due to the wide extent of the impacted area, local levelling campaigns did not cover the full extent of the impacted area. Prior to the reduction in groundwater abstraction, the area was subsiding -1 to -2 mm/yr. Between 2016 to 2019 displacement rates remained fairly stable, with a gradual reduction in subsidence rates from -1 to -2 mm/yr to to 0.0 mm/yr.
However, from June 2019, the ground started to swell locally at +1 mm/yr. Using these swelling observations with InSAR, it was decided to pause the phase out during 2021, prolonging the significant costs of extracting water by another year. At the moment the area is being continuously monitored with InSAR measurements every 11 days, and early 2022 it will be assessed whether the phase out can be continued in 2022.

In semi-arid regions characterized by large agricultural activities, a high volume of water is needed to cover the water requirements for agricultural production. Due to low precipitation and the associated limited availability of surface water, aquifers often represent the main source of irrigation water in these regions. Mostly, the information about the abstraction of the groundwater resources and its management are not well investigated because of technical and financial restrictions. Thus, there is a high demand and need for improved and sustainable monitoring approaches.
Over the past decades, remote sensing has been established as an effective and powerful tool to monitor the planet’s surface. The Copernicus Program of the European Commission (EC) in partnership with the European Space Agency (ESA) offers strong possibilities in satellite based monitoring using remote sensing techniques. Since the first Sentinel mission has been launched in 2014, a solid database of satellite imagery with a high temporal resolution has been made available for everyone based on a free and open data policy. Particularly, Interferometric Synthetic Aperture RADAR (InSAR) techniques have gained a higher focus for groundwater management and may help to assess reliable information on the subsurface.
In this study in the framework of German-Moroccan international cooperation, the Chtouka region with its eponymous aquifer in South Morocco has been chosen. It represents a region with great importance for the export of agricultural products and the national trade balance and, therefore, depends on an anticipatory sustainable ground water management. In addition, high groundwater abstraction rates significantly change the flow dynamics of this coastal aquifer and lead to increased saltwater intrusion deteriorating the groundwater quality in the long term.
Sentinel 1 C-band data has been used in order to measure the ground displacement and the velocities over the past six years. Two smaller areas in the Chtouka region has been identified where interferometric analysis based ground deformation maps and available piezometric head measurements are investigated. Based on these observations, a correlation can be made between the ground motion and the change in groundwater level. The results can improve a sustainable groundwater management by directly quantifying the groundwater abstraction where in-situ data is insufficient and by filling gaps in monitoring data. In addition, simulations can be run to simulate future ground motions to support the regulation of the groundwater abstraction for agricultural purposes.

Land subsidence is a geological hazard, which can be induced by anthropogenic factors, mainly related with the extraction of fluids. The San Luis Potosí metropolitan area has suffered considerable damage induced by the overexploitation of the aquifer-system over the past decades. The city is placed on a tectonic graben delimited by mountain systems. The basin was filled over the years by pyroclastic material and alluvial and lacustrine sediments, which compose the upper aquifer, and the top layer of the deep aquifer. With a semi-arid climate and no permanent watercourse, the population water supply depends on small surrounding dams and groundwater resources. Owing to these conditions, nowadays 84% of the water demand in the valley is covered by groundwater. Consequently, the aquifer static level depletion has fallen up to 95 m from 1970s. The continuous decline increases the effective stress acting on unconsolidated Quaternary sediments, and therefore, the areas with higher accumulated thickness (up to 600 m in the center of the aquifer) consolidate. In this study the relationship between piezometric level evolution and land subsidence is analyzed. To this aim, we applied Coherent Pixels Technique (CPT), a Persistent Scatterer Interferometry (PSI) technique, using 112 Sentinel-1 acquisitions from October 2014 to November 2019 to estimate the distribution of deformation rates. Then, we compared the PSI time-series with the piezometric level changes using 24 wells records for the period 2007-2017. The results indicate a clear relationship between these two factors. The zones with the greatest drawdowns in the piezometric levels match those areas exhibiting the greatest thickness of deformable materials and maximum subsidence. Therefore, the storage coefficient (S) of the aquifer-system was calculated, using the vertical compaction (∆D) measured by means of PS-InSAR data, for a ∆h piezometric level change. The ratio of the change in displacement to the change in ground water level for the continuous and permanent drawdown represent the inelastic storage coefficient (Skv). Skv values obtained from this analysis show an agreement with previous in situ studies, highlighting the usefulness of PS-InSAR derived data to calculate hydrological parameters in detritical aquifers systems affected by land subsidence owing to groundwater withdrawal.

Land subsidence is a geological hazard characterized by the gradual downward movement of the ground surface. It can be induced by natural processes (e.g. tectonic, diagenesis) or human activities (e.g. subsurface fluid extraction). Extensive groundwater withdrawal from aquifer systems is the main factor causing land subsidence in areas where surficial water is scarce. Groundwater pumping causes a pressure decline in the sandy units and in adjacent unconsolidated deposits (aquitards). As a result, the stress exerted by the load of the overlying deposits is transferred to the grain-to-grain contacts increasing the effective intergranular stress. Depending on the compressibility of the soil, the depleted layers (aquifers and intervening aquitards) compact, thus causing land subsidence. Among other risks, compaction permanently reduces the capacity of the aquifer-system to store water. Therefore, assessing land subsidence is a key step to understand and model aquifer deformation and groundwater flow, which can help to design sustainable groundwater management strategies.
Advanced Differential Interferometry Synthetic Aperture Radar (A-DInSAR) is a satellite remote sensing technique widely used to monitor land subsidence. The Sentinel-1 mission, from the Copernicus European Union's Earth Observation Programme, comprises a constellation of two polar-orbiting SAR satellites that provide enhanced revisit frequency and worldwide coverage under a free, full, and open data policy. To handle and process huge and constantly increasing Sentinel-1 archive, the Geohazard Platform (GEP) on-demand web tool initiative was launched in 2015. In this online processing service, SAR images and A-DInSAR algorithms are located together in a friendly interface. The processing chains run automatically in the server with very little user interaction.
The GEP service is particularly useful for preliminary a land subsidence analysis, as all the data and technical resources are external and in addition the processing time is relatively fast. We tested different A-DInSAR algorithms (named Thematic Applications) included in the GEP to explore land subsidence in four water stressed aquifers around the Mediterranean basin. Located in Spain, Italy, Turkey and Jordan, they are characterized by largely different hydrogeologic features. These pilot sites are studied within the framework of the RESERVOIR project that aims to provide new products and services for a sustainable groundwater management model. This project is funded by the PRIMA programme supported by the European Union. The preliminary land subsidence results provided line-of-sight LOS velocity maps obtained from the GEP and allowed us to identify potential deformation over wide areas before carrying out more refined and conclusive A-DInSAR analyses.

Land subsidence triggered by the overexploitation of groundwater in the Alto Guadalentín Basin (Spain) aquifer system poses a significant geological-anthropogenic hazard. In this work, for the first time, we propose a new point cloud differencing methodology to detect land subsidence, based on the multiscale model-to-model cloud comparison (M3C2) algorithm. This method is applied to two airborne LiDAR datasets acquired in 2009 and 2016, both with a density of 0.5 p/m2. The results show vertical deformation rates up to 10 cm/year in the basin during the period from 2009 to 2016, in agreement with the displacement reported in previous studies. Firstly, the iterative closest point (ICP) algorithm is used in the point cloud registration with a very stable and robust performance. LiDAR datasets are affected by several source of errors related to the construction of new buildings and the changes caused by vegetation growth. The errors are removed by means of gradient filtering and cloth simulation filtering (CSF) algorithm. Other sources of error are related to the internal edge connection error in the different flight lines. To address these errors, the smoothing point cloud method by incorporating average, maximum and minimum cell elevation is applied. LiDAR results are compared to the velocity measured by a continuous GNSS station and an InSAR dataset. For the GNSS-LiDAR comparison, a velocity average from a buffer area processed in the cloud point dataset is applied. For the InSAR-LiDAR comparison a 100m*100m grid is computed in order to assess any similarities and discrepancies. The results show a good correlation between the vertical displacement derived from the three different surveying techniques. Furthermore, LiDAR results have been compared with the distribution of soft soil thickness showing a clear relationship. Detected ground subsidence is a consequence of the evolution of the piezometric level of the Alto Guadalentín aquifer system that has been exploited since the 1960s producing a great groundwater level drop. The study underlines the potential of LiDAR to monitor the range and magnitude of vertical deformations in areas that prone to be affected by aquifer-related land subsidence.

Groundwater plays a critical role for ecosystems and is a vital resource for humankind, providing one-third of freshwater demand globally. When groundwater is extracted unsustainably (ie. groundwater extraction exceeds groundwater recharge over extensive areas and for an extended period of time), groundwater levels inevitably decline and can lead to aquifer depletion, which can pose a risk to the sustainability of urban developments and any groundwater-dependent activities. In an ever-changing world, it is increasingly important to effectively manage our aquifer systems to ensure the longevity of the groundwater resources.

This work is undertaken as a partnership between the University of Pavia and the ResEau-Tchad project (www.reseau-tchad.org), with a focus on the urban area of N’Djamena, the fast-growing capital city of Chad. Groundwater contained within the phreatic and semi-confined aquifers underlying the city acts as the main source of water for a population of around one million inhabitants. With an annual growth rate of 7%, the reliance on groundwater for drinking and agricultural purposes is becoming more important. As a result, there is an increasing pressure on urban sanitation infrastructures that have failed to meet the current demand. Additionally, in recent years this area has experienced frequent flooding which is linked to the overflow of the Chari and Logone rivers and increased extreme precipitation events, which may be exacerbated by land subsidence induced by groundwater overexploitation.

Through the use of Advanced Differential Interferometric Synthetic Aperture Radar (A-DInSAR) techniques, land displacement in N’Djamena and the surrounding area has been spatially and temporally quantified for the first time. The current work aims to present two different InSAR processing techniques, Persistent Scatterers (PS) and Small BAseline Subset (SBAS), in a comparative way and also a preliminary analysis of spatial-temporal correlations between deformation measurements and groundwater levels to evaluate a possible cause and effect relationship.

InSAR is a technique that provides a measurement of ground deformation which, in the context of groundwater management, is controlled by the physical parameters of the aquifer such as soil compressibility, thickness, and storativity. Thus, while InSAR results enable large-scale, high resolution measurements of land displacement (on the scale of millimetres), InSAR-derived data itself is not directly quantitative without lithological knowledge of the subsurface. Therefore, the methodology developed to interpret the InSAR data and characterise the groundwater resources of N’Djamena is based on a multidisciplinary approach that integrates limited, in-situ hydrogeological measurements, including groundwater levels collected during a monitoring regime conducted from June 2020 to July 2021, along with the development of a three-dimensional subsurface lithological model based on the collection of available borehole logs and fieldwork validation.

To generate measurements of land displacement, both the PS and SBAS InSAR techniques have been applied to detect surface deformations in N’Djamena and its surrounding area. The PS-InSAR approach analyses interferograms generated with a common master image to produce a signal that remains coherent from one acquisition to another by exploiting temporally stable targets. Alternatively, the SBAS approach relies on small baseline interferograms that maximize the temporal and spatial coherence. In this work, both techniques have been applied in the study area using two time-series of descending and ascending Sentinel-1 Synthetic Aperture Radar images obtained from April 2015 to May 2021. The PS-InSAR technique mainly focuses on the urban area to obtain a high density of PSs, enabling more accurate land deformation measurements. The PS-InSAR vertical deformation rate ranges from -13 mm/yr to 21 mm/yr, while the SBAS values are in the range of -71 mm/yr to 32 mm/yr. The difference in velocity ranges can be explained by the different spatial coverage achieved by the two processing techniques, as the SBAS method provides results even over non-urban areas, which is where the higher displacement rates are estimated. The deformation rate maps obtained from the PS-InSAR and SBAS results are compared from a quantitative and qualitative point of view, taking into account the different types of movement derived from the techniques. The land deformation depicted for the urban area by the two processing techniques indicates a similar pattern of displacement (similar areas of subsidence and uplift). Although the pattern of displacement indicated by the two datasets is similar, the average velocity values obtained with PS-InSAR tend to be noisier than the ones derived using the SBAS technique, particularly when the SBAS time-series shows non-linear deformation trends.

The approach used in this work exploits advanced satellite-based Earth Observation techniques in order to gain further insight into the behaviour of the aquifer system in a region where hydrogeological monitoring is still largely absent. It is anticipated that the findings will help to improve the characterisation of the aquifer and groundwater resource management in the city of N’Djamena and could be further exploited for strategic decisions in sanitation risk management.

Abstract

Water scarcity is a constant concern for millions of people around the world without access to clean water. This reality is also verified in the city of Recife, located in the northeast of Brazil. The municipality is built on an estuarine plain composed of several rivers (Capibaribe, Beberibe, Tejipió). Over the past 50 years, population growth combined with periods of surface water crisis have significantly boosted groundwater use. The capture of this resource, however, occurs in an indiscriminate way in a large part of the city. Groundwater management is inefficient. The biggest limitation is evident in the control of wells, which are estimated at more than 13 thousand. Most are illegal and of unknown existence to inspection bodies. Over the decades, the weakness in groundwater management has contributed to the overexploitation of confined aquifers. The excessive removal of water resources from the subsoil has caused a reduction in the piezometric level to values above 100 m in the southern part of Recife, in the densely built-up neighborhood of Boa Viagem. This implies a strong risk of land subsidence. The geological phenomenon causes surface lowering and causes greater concern in urban areas. The deformation of the terrain can generate relevant impacts on infrastructure and the environment, causing economic and social damage, and compromising people's quality of life. In general, several cities around the world live with this situation. In addition to natural causes, the main occurrences result from human action in the accentuated exploitation of aquifers. The aim of this research is to use interferometry synthetic aperture radar (InSAR) to detect land subsidence in the coastal plain of Recife caused by the exploitation of groundwater resources. The use of this technology is seen as an innovation to current recurrent practices, based on terrestrial measurement techniques. The procedure is performed with a persistent scatterer interferometric (PSI), in the analysis of SAR data with a single-look complex (SLC) processing level, formed by satellite images: COSMO-SkyMed (ascending orbit, HH polarization, X-band), Sentinel -1 (descending orbit, VV polarization, C band) and PAZ (ascending and descending orbit, HH polarization, X band). Preliminary results reveal a correlation between land subsidence and reduction of groundwater in the southern zone due to water desaturation in the neighborhood of Boa Viagem, with a velocity close to -3 mm/year. Thus, the wide availability of interferometric data from satellite SAR missions associated with an advanced processing method should provide a better understanding of processes that generate superficial instability – such as the land subsidence. Using InSAR provides opportunities to test hypotheses and investigate situations that were previously unlikely due to the lack of adequate information. Its application opens the way for new perspectives in the study of the compaction of compressible sediments in the Recife coastal plain as a result of the decline in groundwater levels.

Keywords: land subsidence; groundwater; Recife; SAR interferometry

Cyanobacterial Harmful Algal Blooms are an increasing threat to coastal and inland waters. These blooms can be detected using optical radiometers due to the presence of phycocyanin (PC) pigments. However, the spectral resolution of best- available multispectral sensors limits their ability to diagnostically detect PC in the presence of other photosynthetic pigments. To assess the role of spectral resolution in the determination of PC, a large (N=905) database of co-located in situ radiometric spectra and PC collected from a number of inland waters are employed. We first examine the performance of select widely used Machine Learning (ML) models against that of benchmark algorithms for hyperspectral remote sensing reflectance (Rrs ) spectra resampled to the spectral configuration of the Hyperspectral Imager for the Coastal Ocean (HICO) with a full-width at half-maximum of < 6 nm. The ML algorithms tested include Partial Least Squares (PLS), Support Vector Regression (SVR), eXtreme Gradient Boosting (XGBoost), and Multilayer Perceptron (MLP). Results show that the MLP neural network applied to HICO spectral configurations (median errors < 65%) outperforms other scenarios. This model is subsequently applied to Rrs spectra resampled to the band configuration of existing hyper- (PRecursore IperSpettrale della Missione Applicativa; PRISMA) and multi-spectral (OLCI, MSI, OLI) satellite instruments and of the one proposed for the next Landsat sensor. The performance assessment was conducted for a range of optical water types separately and combined. These results confirm that when developing algorithms applicable to all optical water conditions, the performance of MLP models applied to hyperspectral data surpasses that of those applied to multispectral datasets (with median errors between ~ 73% and 126%). Also, when cyanobacteria are not dominant (PC:Chla is smaller than 1), MLP applied to hyperspectral data outperforms other scenarios. The MLP model applied to OLCI performs best when cyanobacteria are dominant (PC:Chla is equal or greater than 1). Therefore, this study quantifies the MLP performance loss when datasets with lower spectral resolutions are used for PC mapping. Knowing the extent of performance loss, researchers can either employ hyperspectral data at the cost of computational complexity or utilize datasets with reduced spectral capability in the absence of hyperspectral data.

Large and globally representative in situ datasets are critical for the development of globally validated bio-optical algorithms to support comprehensive water quality monitoring and change detection using satellite Earth observation technologies. Such datasets are particularly scarce and geographically fragmented from inland and coastal waters. This is at odds with the importance of these waters for supporting human livelihoods, biodiversity, and cultural and recreational values. These shortcomings create two challenges. The first and major challenge is to collate these datasets and assess their compatibility concerning methodologies used and quality control procedures applied. The second challenge is to identify biases and gaps in the global dataset, in order to better direct future data collection efforts.

Our ongoing effort is to improve the availability of such datasets by providing open access to a large global collection of hyperspectral remote sensing reflectance spectra and concurrently measured Secchi depth, chlorophyll-a (Chla), total suspended solids (TSS), and absorption by colored dissolved organic matter (acdom). This dataset represents an expansion of data originally collated for a collaborative NASA-ESA-led exercise to assess the performance of atmospheric correction processors over inland and coastal waters (ACIX-Aqua). Its suitability for the development of globally applicable algorithms has been demonstrated by its use for developing novel approaches for the retrieval of Chla and TSS concentrations from a range of satellite sensors.

Our dataset contains relevant entries from the commonly used SeaWiFS Bio-optical Archive and Storage System (SeaBASS) and Lake Bio-optical Measurements and Matchup Data for Remote Sensing (LIMNADES) data archives and, in return, contributes thousands of new entries to these and other repositories. It encompasses data from inland and coastal waters distributed across five continents and a comprehensive range of optical water types. Our accompanying biogeographical data analysis contributes to a value-added dataset to aid in the identification of underrepresented geographical locations and optical water types, useful for targeting future data collection efforts.

To ensure the ease of use of this dataset and support the analysis of uncertainties and algorithm development, metadata covering the viewing geometry and environmental conditions were included in addition to hundreds of matched scene IDs for a number of multispectral satellite sensors (e.g. roughly 450 clear-sky match-ups for Landsat 8’s Operational Land Imager (OLI)), making it easier to validate algorithm performance in practical applications.

In curating this dataset, we had to overcome considerable challenges, including technical difficulties, such as variable measurement ranges of instruments, and others due to the fact that the data originated from a community-initiative of multinational researchers working on projects with a diverse range of objectives. Substantial data harmonization efforts to align different instrumentation, field methodologies, and processing routines were needed.

We conclude, our effort was a very worthwhile undertaking as demonstrated by a series of novel contributions and the publication of eight peer-reviewed research articles (at the time of writing). We expect that open access to this dataset will support the development of increasingly data-intensive algorithms for the retrieval of water quality indicators, including those for next-generation hyperspectral satellite sensors, e.g. sensors from the upcoming Surface Biology and Geology (SBG), Environmental Mapping and Analysis Program (EnMap), PRecursore IperSpettrale della Missione Applicativa (PRISMA) Second Generation (PSG), Copernicus Hyperspectral Imaging Mission for the Environment (CHIME), and FLuorescence EXplorer (FLEX) missions. We believe that this will stimulate the discussion of a framework for the future collection of fiducial reference data towards global representativeness.

The objective of this work is to develop new classifications of optical types of water, using remote sensing reflectance (Rrs) measured by satellite as basis. The Rrs of several lakes with different water types are selected and labelled by an expert, identifying each optical water type (OWT) manually. The area of study are different reservoirs and lakes on the eastern Iberian Peninsula, and the Rrs are extracted from Sentinel 2-MSI atmospherically corrected imagery.

The OWT classifiers used here are framed in what is known as supervised classifiers, since they use the previous information given by the user to determine the classes to be detected. In order to classify these data we need atmospherically corrected images, for which the Case 2 C2RCC (Case 2 Regional Coast Color) algorithm developed by Doerffer et al. (2016) and available at SNAP has been applied. The collection of the refectance samples for training and testing has also been carried out using SNAP GUI.

Jupyter Notebooks are in place for the training, testing, application and validation of the models. The classifications generated can help to better understand the seasonal and spatial variations of the studied water masses, being a basic support in the monitoring programs of lakes and reservoirs. It is possible to use the OWT classification as final products to analyse changes in water types related to the different water dynamics of the lakes, or they can considered an intermediate products that could help in the subsequent selection of the water quality data extraction algorithm (for example, chlorophyll concentrations or total suspended matter) generated and adapted to specific types of water (Eleveld et al., 2017, Stelzer et al. 2020).

Results on the classifications tested, and validation of those results, will be analysed and considerations taken about the transfer learning to other lakes in Europe.

Over the last two decades, the primary focus of the development and application of remote sensing algorithms for lake systems was the monitoring and mitigation of eutrophication and the quantification of harmful algae blooms. Oligotrophic and mesotrophic lakes and reservoirs have consequently received far less attention. Yet, these systems constitute 50 – 60% of the global lake and reservoir area, are essential freshwater resources and represent hotspots of biodiversity and endemism.
Uncertainties associated with remote sensing estimates of chlorophyll-𝘢 (chla) concentration in oligotrophic and mesotrophic lakes and reservoirs are typically much higher than in productive inland waters. Uncertainty characterisation of a large 𝘪𝘯 𝘴𝘪𝘵𝘶 dataset (53 lakes and reservoirs: 346 observations; chla < 10 mg/l, dataset median 2.5 mg/l) shows that 17 algorithms, either recently developed or already well established, have substantial shortcomings in retrieval accuracy with logarithmic median absolute percentage differences (MAPD) > 37% and logarithmic mean absolute differences (MAD) > 0.60 mg/l. In the case of most semi-analytical algorithms the chla retrieval uncertainty was mainly determined by phytoplankton absorption and composition. Machine learning chla algorithms showed relatively high sensitivity to light absorption by coloured dissolved organic matter (CDOM) and non-algal pigment particulate absorption (NAP). In contrast, the uncertainties of red/near-infrared (NIR) algorithms, which aim for lower uncertainty in the presence of CDOM and NAP, were linked to the total absorption of phytoplankton at 673 nm and variables related to backscatter. Red/NIR algorithms proved to be insensitive to chla concentrations below 5 mg/l.
Bayesian Neural Networks (BNNs) for OLCI and the Sentinel-2 Multispectral Instrument (MSI) were developed as an alternative approach to specifically address the uncertainties associated with chla concentration retrieval in oligotrophic and mesotrophic inland water conditions (data from > 180 systems, n > 1500). The probabilistic nature of the BNNs enables to learn the uncertainty associated with a chla estimate. The accuracy of the provided uncertainty interval can be consistently improved when as little as 10% of the training data are set aside as a hold-out set. The BNNs improve the chla retrieval when compared versus established and frequently used algorithms in terms of performance over the expected training distribution, when applied to independent regions outside of those included the training set and in the assessment with OLCI and MSI match-ups.

Lakes play a crucial role in the global biogeochemical cycles through the transport, storage and transformation of different biogeochemical compounds. Furthermore, their regulatory service appears to be disproportionately important relative to their small areal extent. The global temperatures are expected to increase further over the coming decades, and economic development is driving significant land-use changes in many regions. Therefore, the need for improved understanding of the interactions between lake biogeochemical properties and catchment characteristics, as well as innovative approaches and techniques to get required high-quality information for large scale has never been greater. Unfortunately, only a tiny fraction of lakes on Earth are observed regularly and data are typically collected at a single point and provide just a snapshot in time. Using remote sensing together with high-frequency buoy measurements is one of the options to mitigate these spatial and temporal limitations. Until very recently, there have been no suitable satellites to perform lake studies on a global scale. The technical issues that were hampering remote sensing of lakes for a long time have been partly solved by the European Space Agency with the launch of Sentinel-2A in 2015 and Sentinel-2B in 2017 (S2). S2 covers the whole world, has a very good radiometric resolution and allows data acquisition at 10 m and 20 m resolution, which permits assessment of an unprecedented number of lakes globally. Still, remote sensing products of lakes have been rarely validated and often with poor results. The main problem is a lack of in situ data, which are needed for validating and improving remote sensing products. Using the high-frequency buoy measurements might be the solution. It will enable the validation of remote sensing products more accurately as it increases the probability of getting match-up data. Therefore, combining S2 capabilities, high-frequency measurements and conventional sampling data we firstly aim to estimate the biogeochemical properties (coloured dissolved organic matter, chlorophyll a, total suspended matter, primary production, dissolved organic carbon, total phosphorus and total nitrogen) in optically different European lakes to test which of the biogeochemical properties may be successfully estimated from the S2 data. Secondly, combining remote sensing capabilities with the increasing potential of Geographic Information System and land cover maps we aim to study the interactions between lakes biogeochemical properties, meteorological factors and catchment characteristics with high accuracy in large scales. The expected results will improve our understanding of the role of lakes in the global biogeochemical cycles and have a strong applied impact allowing to make reliable recommendations for decision-makers and lake managers for different ecological, water quality, climate and carbon cycle applications, and improving significantly the cost-efficiency of lake monitoring both regionally and globally.

In lake rich regions, protecting water quality is critically important because of the ecological and economic importance of recreational activities and tourism. To ensure the health of inland aquatic ecosystems on both a local and regional scale, more comprehensive monitoring techniques to complement conventional field sampling methodologies are needed for effective management. For over 25 years, our previous statewide water quality mapping in Minnesota, USA has primarily relied on Landsat satellites. However, measurements have been limited to water clarity and colored dissolved organic matter (CDOM) due to inherent Landsat sensor spectral band configurations. Sentinel-2 Multispectral Imager (S2/MSI) on the other hand offers several red-edge bands that increase the accuracy of chlorophyll concentrations. The increased temporal coverage of S2/MSI along with Landsat-8 Operational Land Imager L8/OLI and recently launched Landsat-9 L9/OLI-2 enables more frequent monitoring of Earth’s inland water bodies and permits routine mapping of water quality parameters.

To utilize these capabilities, we have developed field-validated methods and implemented S2/MSI and L8/OLI image processing techniques in an automated pipeline built in a high-performance computing environment that generates Level-3 (L-3) satellite data products for lake water quality monitoring and management. Machine-to-machine access to ESA Copernicus and U.S. Geological Survey servers allows for the synergistic acquisition of L-1 S2/MSI and L8/OLI imagery to supply the demand for near-real time data. Newly acquired imagery can be immediately sent through multiple scripted processing modules, which include (1) identifying and omitting potentially contaminated pixels caused by clouds, cloud shadow, atmospheric haze, wildfire smoke and specular reflection, and (2) classification of water pixels through a normalized difference water index (nNDWI) to delineate a scene specific water mask. The combined masks result in qualified pixels which advance to (3) a modified SWIR-based aerosol atmospheric correction to the retrieval of remote sensing reflectances (Rrs). The atmospheric correction produces a harmonized reflectance product between S2/MSI and L8/OLI pixels in which modeled L-3 type water quality data products are derived. Calibrated L-3 water quality models including water clarity, CDOM, and chlorophyll-a, rely heavily on field validated datasets to account for the dynamics of optically complex lake systems of the region. To this extent, sampling efforts in the summer months constrain uncertainties between satellite-derived and surface water properties caused by varying atmospheric conditions and calibrate/validate water quality retrieval algorithms to yield verifiable water products. As new field validation data become available at season-end, scripted modules within the processing chain can be modified accordingly and applied to incoming and previously processed imagery if any resulting water quality product models need improvement. Finally, the data can be made available to the public in an online map viewer linked to a spatial database that allow for statistical summaries at different delineations and time windows, temporal analysis and visualization of water quality variables. The Minnesota LakeBrowser (https://lakes.rs.umn.edu/) provides an example of the data that is being produced through this project. Due to the cloud cover in the Midwest, we determined that monthly open water (May through October) pixel level mosaics work best for statewide coverage. Lake level data is determined for each clear image occurrence and compiled in csv files that can be used to calculate water quality variables for different timeframes (e.g. monthly, summer (June-Sept)) and linked to a lake polygon layer that can be used for geospatial analysis and included in a web map interface. For Minnesota the lake level (2017-2020) data includes 603,678 daily lake measurements of chlorophyll, clarity and CDOM (1,811,034 total) and will be updated on a regular basis.

This unique data source dramatically improve data-driven resource management decisions and will help inform agencies about evolving water quality conditions statewide. In terms of decision-making, the production of frequent, near-real time data on water clarity, chlorophyll-a, and CDOM across large regions can enable water quality and fisheries managers to better understand lake ecosystems. The improved understanding will yield societal benefits by helping managers identify the most effective strategies to protect water quality and improve models for increased fisheries production.

The Finnish Environmental Administration has invested in and advanced the utilization of satellite observations to collect environmental data focusing on water quality. The Copernicus program, along with NASA Landsat-programme, provides long-term opportunities and perspective for this. Starting from 2017, Finnish Environment institute (SYKE) has been developing a publicly open TARKKA (https://syke.fi/TARKKA/en ) web map service through which users can utilize satellite observations. The TARKKA service focuses on providing water quality material and information for status assessment of Finnish water bodies. The need for the water quality monitoring via EO is heavily motivated by the extensive obligations of the EU directives (WFD*, but also MSFD**), the assessment of the state of the Baltic Sea (HELCOM*** holistic assessment, HOLAS), and the assessment of the impact of water protection measures. In Finland, the obligations set by EU for WFD reporting concern about 4500 lake and more than 250 coastal water bodies. As part of SYKE’s water quality EO development, a project named CorEO is working on diversifying and enhancing the water quality service based on satellite observations by introducing new analysis methods based on e.g., artificial intelligence. The improvements bring more user-orientation and visuality in TARKKA service.

In addition to the open TARKKA service, useful data on Finnish waters is also collected in a database available via STATUS interface, directed for the authorities responsible for directive reporting. The EO information database covers most of the water areas or bodies covered by the directives (especially the WFD). Although up to 70% of the satellite observations over Finland are partly cloudy, the database accumulates millions of observations from Finnish water areas every year. During the 3rd round of WFD reporting in 2019, Finnish authorities responsible for the lakes water bodies status assessment utilized EO as one source of information and found it to be beneficial to meet the requirements set by the directive. Approximately 40% of Finnish lake water bodies with WFD reporting obligations are included in the STATUS database. Finnish lakes represent wide range of optically complex waters – many of them are absorption dominated humic waters that form one extreme part of Case II waters. After the reporting, the database has been utilized to provide automated information on water quality and has been linked to various services providing information for citizens and authorities e.g. Marine Finland (https://www.marinefinland.fi/en-US/The_Baltic_Sea_now).
In the spring of 2020, automatic production of satellite observations was introduced, and proved to work fluently during the Covid-19 era; the processing, quality assurance and distribution of satellite observations went on schedule. As a side-result, the use of the TARKKA web service increased significantly during the spring and summer. Currently, the main challenge in data production is the vast and growing mass of observations as well as the development of related archiving and computing capacity to meet the future needs over the next ten years. For the following years, the development work will focus on improving the information content of existing services to become more user oriented. This includes development of methods that bring up the relevant part of the information related to water quality in various parts of the lakes in Finland from the vast amount of satellite observations. One of the first demonstrations for this was a service providing lake-specific information on cyanobacteria blooms over 43 Finnish lake districts in the summer 2021. For each of the lake district, the service provided also historical datasets as a background information, dating back to year 2013. From the user's point of view, it is useful to highlight the observations that illustrate the state of the areas requiring more attention or intensive monitoring. Recent development enhances the monitoring and surveillance of the state of the lakes based on satellite observations combined with other types of observation, like station water sampling and automated station observations.
In particular, the development focuses on the visuality and communicability of observations. One of the focus points is the development of automatic detection of sudden and long-term changes and identifying problem areas that require special attention (including nutrient sources, coastal estuaries, cyanobacteria). Anomaly tracking using artificial intelligence is another focus area for the development. In most cases, the spatial resolution of Sentinel-2-series MSI and Landsat-series OLI are sufficient to capture and identify the user needs like river water impact areas (turbidity and humus interlinked with nutrients), large and medium dredging areas, nuclear power plant condensate temperatures (TIRS-instrument), coastal, lake and offshore algae. The solutions enhance the introduction of data suitable for environmental monitoring in Finland.

*WDF: water Framework Directive, ** MSFD = Marine Strategy Framework Directive, ***HELCOM = Helsinki Commission, i.e. Baltic Marine Environment Protection Commission

Cyanobacteria are successfully growing in many waterbodies, causing potentially toxic surface blooms, hampering the recreational activities, impeding the water usage and causing problems for lake biota. Lake Peipsi is the largest transboundary waterbody in Europe, which consists of three parts: Lake Peipsi s.s., Lämmijärv and Lake Pihkva. Naturally occurring cyanobacterial blooms are a characteristical feature for this eutrophic lake, dominated by Gloeotrichia echinulata, Aphanizomenon, Dolichospermum and lately Microcystis with increasing abundance, especially in L. Lämmijärv. Regular national in situ monitoring covers Estonian side of the lake once per month from minimum 7 locations during vegetation period, but with in situ methods is complicated to give an overview of the bloom dynamics, its onset, the length of the bloom presence and its spatial extent. Remote sensing methods give complement information more frequently and allow better overview to the bloom on spatial scale. We used Sentinel 3 A and B/OLCI FR L1 images with MCI and regional conversion factors for Chlorophyll a (Chl a) concentration assessment, whereas time period of 2016-2021 was studied. Chl a values in Peipsi s.s. were generally lower (below 40 µg/L) in comparison to Lake Lämmijärv and Lake Pihkva, where higher values were present (> 75 µg/L and > 100 µg/L, respectively) during 2019-2021.
The threshold for cyanobacterial bloom presence/absence may be difficult to set, for example according to World Health Organisation the bloom starts already from 10 µg/L of Chl a, but for Peipsi this is not suitable, since majority of values measured during vegetation period are higher. As a lake-specific solution the presence of cyanobacterial blooms was assessed via taking lake-part specific long-term median value of Chl a from historical records (1984-2015) of in situ measurements for the period of June to September + 5%. Cyanobacterial bloom duration and extent differed in lake parts and between different years. Bloom generally started in Peipsi s.s. earlier than in other lake parts, and bloom duration was there longest, lasting >100 days with maximal coverage 68±19 % of the total lake area. Cyanobacterial concentration was higher in Lämmijärv, during the maximum extent of the bloom Lämmijärv was nearly entirely covered by cyanobacteria, with the exception of 2018, when coverage remained below 76%. During 2018 bloom coverage was also lowest in L. Pihkva ( < 30%). In general, the bloom duration in L. Pihkva was similar or shorter than in Lämmijärv, but with higher cyanobacterial biomass.

Atmospheric correction over inland and coastal waters is one of the major remaining challenges in aquatic remote sensing, often hindering the quantitative retrieval of biogeochemical variables and analysis of their spatial and temporal variability within aquatic environments. The Atmospheric Correction Intercomparison Exercise (ACIX-Aqua), a joint NASA – ESA activity, was initiated to enable a thorough evaluation of eight state-of-the-art atmospheric correction (AC) processors available for Landsat-8 and Sentinel-2 data processing. Over 1000 radiometric matchups from both freshwaters (rivers, lakes, reservoirs) and coastal waters were utilized to examine the quality of derived aquatic reflectances (ρ ̂_w). This dataset originated from two sources: Data gathered from the international scientific community (henceforth called Community Validation Database, CVD), which captured predominantly inland water observations, and the Ocean Color component of AERONET measurements (AERONET-OC), representing primarily coastal ocean environments. The volume of our data permitted the evaluation of the AC processors individually (using all the matchups) and comparatively (across seven different Optical Water Types, OWTs) using common matchups. We found that the performance of the AC processors differed for CVD and AERONET-OC matchups, likely reflecting inherent variability in aquatic and atmospheric properties between the two datasets. For the former, the median errors in ρ ̂_w (560) and ρ ̂_w (664) were found to range from 20 to 30% for best-performing processors. Using the AERONET-OC matchups, our performance assessments showed that median errors within the 15 – 30% range in these spectral bands may be achieved. The largest uncertainties were associated with the blue bands (25 to 60%) for best-performing processors considering both CVD and AERONET-OC assessments. We further assessed uncertainty propagation to the downstream products such as near-surface concertation of chlorophyll-a (Chla) and Total Suspended Solids (TSS). Using satellite matchups from the CVD along with in situ Chla and TSS, we found that 20 – 30% uncertainties in ρ ̂_w (490≤λ≤743 nm) yielded 25 – 70% uncertainties in derived Chla and TSS products for top-performing AC processors. We summarize our results using performance matrices guiding the satellite user community through the OWT-specific relative performance of AC processors. Our analysis stresses the need for better representation of aerosols, especially absorbing ones, and improvements in corrections for sky- (or sun-) glint and adjacency effects, in order to achieve higher quality downstream products in freshwater and coastal ecosystems.

AIM INTRODUCTION
Monitoring water quality is valuable since the changes that may occur in water bodies have severe socio-economic and environmental impacts. Such an influence is evident in Timsah lake, the biggest water body of Ismailia district in Egypt, which has been the objective of this research. The main aim of this research is to estimate the changes in the water quality of the area during the period of 2014-2020. Timsah Lake was subjected to significant environmental pressures, caused by various anthropogenic activities in Ismailia city. From satellite observations in the optical part of the spectrum, we can retrieve the concentrations of different constituents (pure water, chlorophyll, sediments, coloured dissolved organic matter) and also, we can use the satellite data to detect changes in the zone surrounding the water bodies.
Within the framework of increasing world trade, increases in the size of ships, and the need of the Egyptian economy to develop its resources, it was imperative to expand the current Suez Canal to cope with the increasing future world trade (EEAA, 2014). A new canal was implemented on the 5th of August 2014 parallel to the existing one; Suez Canal (SC). It is suspected that quality changes may be arisen due to the construction of the New Suez Canal (NSC). Timsah Lake has a strategic location on the Suez Canal, which is the main route joining the continent of Africa with Asia and Europe the evaluation of human activities, basically encompassing navigation as a pathway for trading ships with other countries, fishing which provides a vital source of food and income for local population and tourism.
DATA- METHODOLOGY
In order to achieve the goal of this study, free satellite images from Landsat 8 OLI and Sentinel-2 satellites have been exploited to analyse the objective of this research study. To be more specific, for Landsat 8 the Level-1 and Level-2 scenes were obtained from the United States Geological Survey (https://earthexplorer.usgs.gov). Landsat 8 carries the Operational Land Imager (OLI) and the Thermal Infrared Sensor (TIRS) instruments.
In addition, Sentinel-2 is a European wide-swath, high-resolution, optical multi-spectral imaging mission. The full mission consists of the twin satellites flying in the same orbit but phased at 180°, is designed to give a high revisit time of 5 days and has 13 spectral bands (VIS, NIR, SWIR) (Drusch, et al., 2012). The Multispectral Instrument (MSI) Level-1C (L1C) Sentinel-2 scenes are available from the open-source of Copernicus Open Access Hub (https://scihub.copernicus.eu). Finally, the processing of the data has been carried out by the free and open ESA’s SNAP, the Harris Geospatial Solution’s ENVI and the final maps have been exported with the ESRI’s ArcGIS software.
As far as the methodologies are concerned, Principal Component Analysis (PCA) and Case 2 Regional Coast Colour (C2RCC) algorithms were applied for the purpose of monitoring the physical properties of different water characteristics encompassing pure water, chlorophyll-a, sediments, and Total Suspended Matter (TSM). In order to detect changes in the area of Ismailia city including the Suez Canal proximate area, Landsat 8 images have been used and the PCA analysis has been applied. PCA transforms an original correlated dataset into a substantially smaller set of uncorrelated variables that represent most of the information presented in the original dataset (Richards 1994; Jensen 2005). For the generation of the image-components we have adopted the PCA method by using the Landsat 8 L2 products dated 22/08/2014 and 06/08/2020 the visible and near-infrared bands, in total four bands from each date. The correlated variables (original bands) are transformed into other uncorrelated variables (principal components images). The maximum original information contained by these, with a physical meaning that needs to be explored. It is also proven, that the first three principal components may contain more than 90 percent of the information in the original seven bands.
Concerning the C2RCC, the objective of this algorithm is to determine optical properties and the concentrations of constituents of case 2 waters. Case 2 water is defined as a natural water body, which contains more than 1 component of water constituents, which determine the variability of the spectrum of the water leaving radiance and is presented in coastal seas, estuaries, lagoons and inland waters (Morel & Prieur, 1977; Gitelson et al. 2007). For this study, 7 Sentinel 2 A & B, Level 1C products were used and acquired by the satellite during August month each year from 2015 to 2020. Also, Landsat 8 Level 1 products, for month August 2014 and 2020 have been processed using the C2RCC. The processing of the Sentinel 2 and Landsat 8 images could be divided into two steps. The first one is the pre-processing including resampling (in this case resample in 40 m/pixel) so that all bands have the same spatial resolution and then is the subset of the image over the study area. The second one concerns application of the C2RCC algorithm in order to detect the amount of seawater suspended matters and chlorophyll-a and also perform the atmospheric correction. The final product is a thematic facilitating the knowledge content the unit is cubic gm-3 which is unit of volume cubic gigametre. Finally, the data were exported in GeoTIFF format and then imported into ArcGIS software, where the final maps are produced.
RESULTS -DISCUSSION
The results of the present work and the figures that emerged revealed that the original hypothesis of this research was proven correct, which presumed that the main source of pollution was the NSC. The obtained values of the different water characteristics that the Western Lagoon -located to the Western part of Timsah lake and connected with it through one inlet in the western side- and its emerging streams (Abu Atwa drain) are considered as contamination starting points. Generally, TSM and Chlorophyll-a results from Sentinel-2 data, indicate that during August 2015 there are lower values of TSM and Chlorophyll-a, with a range between 4 and 17 g m-3 and 2 and 11 g m-3, respectively. While the higher values of the TSM and Chlorophyll-a concentration, appears during August 2018 for the TSM, around 23 and 50 g m-3 and for Chlorophyll-a during August 2016 and 2018, around 20 and 40 g m-3. Specifically, high levels of TSM are evident and concentrated in the Western Lagoon. These values are at their pick especially during August 2018, 2019 and 2020. Regarding the Landsat-8 results, the TSM in August 2014 is lower than the following periods until 2020, where higher values are attributed mainly to the Western Lagoon and the inlet of the Timsah lake. For Chlorophyll-a, the highest value is recorded in 2014 than any other year, concentrated in NSC and the Western Lagoon.
Concerning the results of PCA, from the analysis of eigenvalues and eigenvectors, in combination with the interpretation of the principal components, the component images which are suitable in order to perceive the possible spatial changes, are selected. Moreover, the first component (PC1) corresponds to the brightness image (information concerning topography and albedo) and contains 78.35% of the information. The second component (PC2) calculates the spectral information related to all the transformations that took place during the period 2014 to 2020 and contains 11.12% of the information. There is the difference image between the two dates, resulting from the negative contribution of the original spectral bands of the first date (2014) and the positive contribution of the original spectral bands of the second date (2020). The last PC images contain a small amount of information relative to other applications and “noise” (Psomiadis et al. 2005).
To conclude, this study demonstrates a correlation between the results and the overall change in the area, since human activity and technological development is increasing. The contaminants of aquatic and wetland environments are a common situation due to the changes in the surrounding area. Such changes are tourism, agriculture, hotels, leisure infrastructures, which are built throughout the last years. As is confirmed from El-Serehy et al. (2018) and Abd El-Azim et al. (2018), there are three major sources responsible for water quality changes in Timsah Lake (area of interest) such as agricultural drainage, anthropogenic activities, untreated domestic and industrial waste discharges. Due to the lack of in situ data, the results cannot thoroughly confirm this hypothesis.
In line with the main goal, the recorded values of the TSM have shown that in the past years the connection between the Western Lagoon, the Abu Atwa Drain, Ismailia Canal and the Timsah lake appears to be the contamination sources that degrade the quality of water. This might reveal that Lake Timsah is a high eutrophic lake, as it was also pointed out by Mehanna et al. (2016). In our future work, we would like to add in situ data to prove our hypothesis.

Global warming affects ecosystems worldwide. Among others, climate change can trigger lake ecosystems shifts. Increasing trends for lake water temperature have been suggested from single case studies but also at the global scale. Increasing water temperature intensifies the thermal stratification of deep lakes, reducing the intensity of vertical mixing. Warming will thus likely alter the mixing regime of lakes substantially this century, as suggested by recent lake model simulations. This transition potentially leads to abrupt shifts in global lake ecosystems.
A reduction in mixing intensity and frequency can have severe implications for the entire lake ecosystem. For example, reduced deep water renewal hinders the vertical transport of oxygen from epilimnion to hypolimnion and can increase the extent and duration of seasonal hypoxia (low oxygen). Contrarywise, stratification suppresses nutrient resupply from the deep water to the surface layer. Both affect lake primary productivity and its entire food web. Records to characterize mixing regime anomalies and ecosystem shifts or their underlying mechanisms are scarce because they require long and dense time series, requirements often not met by traditional monitoring records.
We review and synthesize information on the detection of regime shifts in lakes worldwide. We suggest three main sources of data that can be used to detect lake ecosystem shifts: sediment coring, high-frequency in-situ measurements, and remote sensing. Remote sensing data started to be used for this purpose in a later stage but its potential to monitor several lakes at the same time allows studying a wider range of lakes. Our synthesis of the literature for more than 700 studies of lake regime shifts shows that to date remotely sensed time series of lake surface water temperatures (LSWT) have been mainly based on spatial averages of lake surface water temperature, neglecting the spatial dimensions of global LSWT products. However, the horizontal gradients could help a better understanding of the internal processes of lakes and the identification of lake mixing or ecosystem anomalies.
Seasonal overturning often occurs at different times across the lake. Thus, the spatial character of remotely sensed data can reveal important processes in freshwater systems and can help assess the long-term variability in the overturning behavior of large lakes in the context of climate change. However, limnologists have so far not extensively explored the spatially distributed character of remotely sensed data. We aim at developing a methodology to detect anomalies or shifts of lake ecosystems by using the spatial patterns of remotely sensed lake water properties (LSWT and ecological variables like turbidity and chlorophyll), and link such patterns to documented anomalies or shifts of lake ecosystems. Here, we exploit the CCI Lakes database from the standpoint of a limnologist, and with an advanced understanding of thermal forcing and ecosystem responses in lakes.

High-mountain lakes are among the most vulnerable ecosystems to climate change, particularly in Mediterranean climates. The Mediterranean region is suffering from an exacerbated rate of climate change compared to the global trends, being considered as a climate change hotspot in the world. The characteristic summer droughts of the Mediterranean climate are being intensified due to the increase in the mean annual air temperature, specially during the summer months, and the decrease in annual rainfall. As a whole, it may produce a decline in the snow accumulation in those regions and an earlier melting of the snow that could eventually affect the hydrology of the ecosystems, among other affected processes. Moreover, high-mountain ecosystems are experiencing elevation-dependent warming, whereby the warming rate is amplified with altitude.
Sierra Nevada National Park (Granada, Spain) is the southernmost mountain range in Europe and constitutes a biodiversity hotspot. In Sierra Nevada there are around 50 small glacially-formed lakes at an elevation of 2800-3100 m above sea level. They are small (surface ranges between 0.01-2.1 ha) and shallow (depth ranges between 0.3-8 m) oligo- to meso-oligotrophic lakes. Because of the high sensitivity of remote lakes to environmental changes, they are considered as excellent sentinels of climate change.
These lakes are hardly accessible. Hence, continuous and regular monitoring of these water bodies is difficult, but essential for their correct management. In this study we will mainly focus on chlorophyll-a (chl-a) since it is a key indicator of phytoplankton biomass and water quality. Remote sensing techniques represent an alternative to the current field samplings that are not always possible or not as frequently as desirable. One of the satellites with the greatest potential is the Multispectral Instrument (MSI) Sentinel-2, due to its high spatio-temporal resolution. Its spatial resolution of up to 10 m may allow the analysis of small waterbodies, and its revisit time of five days might allow the characterisation of temporal dynamics. However, a lack of data for such small lakes as ours has been detected, despite being very frequent ecosystems.
Hence, the aims of this work are (a) to explore the potential of remote sensing techniques using the Sentinel-2 imagery to estimate water quality parameters, mainly chlorophyll, in shallow and small high-mountain lakes in a regional scale and (b) to develop a chl-a estimate model as a tool for eutrophication monitoring since it may increase as a consequence of climate change. This may, in turn, allow the characterization of the lakes in terms of susceptibility to eutrophication since each lake is affected differently by variables such as livestock pressure, tourism, Sahara dust deposition and lake morphometry and watershed features. Finally, (c) this model is intended to be used by the management of the Sierra Nevada National Park and take necessary measures in order to maintain a good ecological status of the lakes.
Achieving our objectives would represent a major breakthrough since until now Sentinel-2 imagery has only been used for this purpose on lakes much larger and deeper than the small high-mountain lakes of Sierra Nevada. This work might represent a baseline for further remote studies of similar ecosystems.
The Sentinel-2 images were obtained and processed through Google Earth Engine (GEE). A first approach was made to select lakes with water-pure pixels. Sseven lakes met this requirement: Río Seco Lake, Yeguas Lake, Caldera Lake, Larga Lake, Mosca Lake, Vacares Lake and Caballo Lake.
Field sampling campaigns were conducted during the ice-free period of 2020 and 2021 obtaining 8 and 40 samples, respectively. An optimal time gap of ±3 days and a maximal of ±5 days were established between the in-situ measurements and the satellite overpass. In each lake an integrated water sample of 1.2 m was collected at a point in which the adjacency and bottom effects were minimized. The samples were stored in dark conditions until we arrived at the laboratory where the chl-a concentration, colored dissolved organic matter (CDOM) and total suspended solids (TSS) were analyzed. Chl-a and CDOM were determined through the filtration of the water samples using pre-combusted Whatman GF/F. Chl-a concentration was assessed by pigment extraction from the filter using ethanol and it was spectrophotometrically analysed. CDOM was determined spectrophotometrically from the filtered water. Finally, TSS were determined from the pre- and post-filtering Whatman GF/F filter weight.
Around 1500 papers relating chl-a and Sentinel-2 published until October 2021 were reviewed. It is worth noting that none of them was conducted specifically in high-mountain lakes. We selected and tested on Sierra Nevada the potential models that had already shown a good performance in oligo- and mesotrophic waters like ours. Traditional empirical, semi-analytical and novel machine learning models were tested. The use of machine learning, a type of artificial intelligence, is increasing in several scientific branches and it is in constant evolution. Hence, it represents a novel approach to chl-a retrieval and an increasing number of papers are showing its high performance on this purpose. According to the literature, the chl-a models that have performed best in waters with similar characteristics to ours use the red band (665 nm), red edge band 1 (705 nm) and red edge band 2 (740 nm). Some of these models are 2BDA (Moses, 2009), 3BDA (Gitelson, 2009), MCI (Gower, 2005) and Toming (2016), among others. However, the model FLH (Fluorescence Light Height) (Buma and Lee, 2020) has shown its high performance in similar waters and it uses the blue band (490 nm), green band (560 nm) and red band (665 nm). Finally, different atmospheric correction algorithms previously published in the literature were tested in combination with the developed chl-a models. Regarding the atmospheric algorithms cited in the bibliography, the algorithms that perform best in clear waters like ours are Polymer and iCor. The last one is the only one that includes correction for the adjacency effect, which is almost unavoidable in our study area. By contrast, the algorithm C2RCC has been shown to fail in presence of adjacency effect.
Taking into account the limited chl-a data taken in-situ during 2020, the model that performed best in our study area was FLH with a determination coefficient R2=0.59. By introducing the data collected during the 2021 field-sampling campaign and increasing the number of experimental replications as well as the number of sampled lakes it is expected to obtain a more accurate model for our study area. Moreover, more pre-existing models developed during 2021 will be tested and different atmospheric corrections will be introduced.

Monitoring of surface water quality is regulated by many national and European regulations and an important aspect of protecting aquatic ecosystems, achieving sustainable development goals, and supporting human well-being. Classical monitoring strategies target in situ monitoring and are time-consuming, cost-intensive and require well-trained personnel and sophisticated analytical labs. Remote sensing techniques can support and extend these monitoring efforts without raising costs and efforts dramatically. Indeed, many research projects convincingly documented that remote sensing is able to assess important water quality variables like turbidity, transparency, chlorophyll, humic substances, water temperature or cyanobacteria. In this context it is astonishing to note that remote sensing does not play a bigger role in governmental monitoring programmes and are hardly used by water managers, state authorities or communes. Why do these institutions do not make more use of the multiple opportunities provided by satellite observations and exploit data providing infrastructures like Copernicus Services, EO-Bowsers or institutional facilities in their governmental tasks?
In this talk we want to identify, explore, and discuss a multitude of reasons that may explain the discrepancy between the rich potential of remote sensing techniques for detecting inland water quality and their limited public utilization. We found a mixture of reasons that often act in concert and, for example, refer to lacking legislative framing, unknown transferability of methods between different kinds of water bodies, missing training and competences in authority staff, lacking harmonisation among states and countries, or the interpretation of remotely sensed data given their more complex data structures. We gained insights into these limitations from communications with German water authorities and European institutions as well as intense discussions with water experts.
We conclude by proposing a structured approach that helps authorities to use remote sensing products in their daily business. This approach includes a sound scientific basis of remote sensing products and harmonized procedures to implement remote sensing into governmental practices. We further report our experiences from co-developing this approach together with state agencies, reservoir authorities and other water-related institutions. This work is embedded in the Copernicus Programme via a research project funded by the German Federal Ministry of Transport and Digital Infrastructure. For further information refer to the webpage of the BIGFE-Project (https://www.ufz.de/bigfe/index.php?de=48596).

AquaWatch Australia intends to integrate Earth Observation (EO) and in situ sensors through an Internet of Things (IoT) connectivity to monitor and predict inland and coastal water quality and habitat condition for a wide range of uses in Australia and across the globe. AquaWatch Australia is designed to measure key aquatic environmental and biogeochemical variables required to understand processes affecting the quality of aquatic ecosystems to provide early warning of extreme events, accurate information on recovering or threatened ecosystems, and to help predict and manage water quality threats by empowering timely management decisions.

AquaWatch Australia is co-led by CSIRO (Australia’s national science agency), the SmartSat Cooperative Research Centre (SmartSat CRC), and other national and international partners, leveraging their longstanding expertise in EO, in situ sensing and modelling of water quality. After delivering an initial concept study (Phase-0), AquaWatch Australia has entered Phase A, including the consolidation of its end-user requirements. The plan is to translate these into systems requirements before entering a production phase, provided sustainable government funding is secured for the next development phases.

AquaWatch Australia was inspired by and follows many recommendations from the CEOS (2018) and IOCCG 2018 reports on Earth observation of Aquatic Ecosystems. The United Nations Sustainable Development Goals (UNSDGs) framework, through its Goal 6 (“Ensure availability and sustainable management of water and sanitation for all”), explicitly highlights the urgent necessity of securing better global access to clean water and its efficient use. Management of freshwater resources is a critical global issue and vital for landscapes, ecosystem functioning, biodiversity, agriculture, and communities.

One of the key proposals of AquaWatch Australia is to develop sovereign Australian capability to build, launch and operate a constellation of satellites optimised for monitoring aquatic systems. With the exception of the coarse spatial resolution GLIMR (a geostationary America’s located satellite), and PACE (for ocean-coastal to large inland waters system), most finer spatial resolution hyperspectral satellite missions including EnMAP, DESIS, PRISMA, SBG, and CHIME have primarily been designed to monitor the land. The sensors and satellites built for AquaWatch Australia will overcome some of the limitations of these sensor systems in terms of the necessary spectral bands, spatial resolution and, if possible, revisit times, which restrict the range of water quality and benthic parameters that can be measured from space as well as the number of waterbodies from which these parameters can be measured. The AquaWatch Australia system is designed to be highly adaptable and can use existing satellite datasets such as Sentinel 2, Landsat 8 and 9, as well as all relevant planned missions to enhance its final output. Combining data from these satellite sensors with the aquatic ecosystem-specific data from AquaWatch could provide opportunities for higher resolution water quality products, using data fusion or blending approaches as required by end-users.

AquaWatch Australia is currently establishing strategic partnerships and has invested in establishing a network of national and international pilot sites. Engaging with water quality researchers and end users across different regions in the world will enable the mission to expand the range of water quality conditions that can be accurately measured and predicted, demonstrate locally-applicable solutions and diversify use cases, and develop local capacity in EO monitoring and prediction to support national and global sustainability agendas. AquaWatch Australia is also intended as Australia's contribution to GEO-AquaWatch.

Phytoplankton and its most common pigment chlorophyll a (Chl a) are important parameters in characterizing lake ecosystems. We compared six methods to detect Chl a in two optically different lakes in boreal region: stratified clear-water Lake Saadjärv and non-stratified turbid Lake Võrtsjärv. Chl a was measured: in vitro with a spectrophotometer and high-performance liquid chromatography; in situ with automated high-frequency measuring (AHFM) buoys as fluorescence, and with a high-frequency optical hyperspectral above-water radiometer (WISPStation); and with various algorithms, applied on data from satellites Sentinel-3 OLCI and Sentinel-2 MSI.

The agreement between all the methods was from weak (R2=0.1) to strong (R2=0.96), while the consistency was better in turbid lake compared to the clear-water lake where the vertical and temporal variability of the Chl a was larger. The agreement between the methods depends on multiple factors. The radiometric measurements are highly dependent on the environmental and illumination conditions, resulting in higher variability in the recorded signal towards autumn. The effect of non-photochemical quenching (NPQ) correction increases with increased PAR and also is highly depended on the underwater light level, which resulted in up to 15% change in the chlorophyll fluorescence in more turbid conditions compared to 81% in clear water lake Saadjärv. Additionally, calibration datasets and applied correction methods required to account for the variability within phytoplankton amount and composition together with the background turbidity also had an effect on the the consistency of the final Chl a estimation.

Synergistic use of data from various sources allows to get a complex overview about a lake in horizontal and vertical scale but prior to merging the data, the method-based factors should be accounted for. These factors can have high impact on the results and lead to poor management decisions while switching approaches to analyze the Chl a patterns e.g. extending time series for estimating the status of the water body based on Chl a according to EU Water Framework Directive.

Water quality remote sensing is increasingly used in an operational context, and several studies in particular for perialpine lakes showed how hydrodynamic modeling can greatly improve the utility of remotely sensed products. Conversely, remotely sensed products can help to improve the performance of hydrodynamic models as a source of dynamic input data, by means of data assimilation, or for validation. With such an interdisciplinary integration of Earth observation techniques, we can take advantage of the forecasting capabilities of data-driven hydrodynamic lake modeling and the synoptic coverage, as well as a regular sampling of high-resolution satellite imagery, i.e., from Sentinel-2.
A first, operational framework that partially established the integrated usage of Earth observation data for Lake Geneva resulted from the ESA project CORESIM (www.meteolakes.ch). As part of the ESA Regional Initiative for the Alpine Region, the project AlpLakes aims to extend this framework functionally and spatially. The two main objectives of AlpLakes are to integrate Sentinel-2 transparency products in hydrodynamic models in order to improve their performance, and to update the models with a particle tracking module for validation with Total Suspended Matter (TSM) estimates from Sentinel-2 data. Ultimately, the project aims at understanding the short- and long-term evolution of the dynamics of freshwater systems with a particular focus on altitudinal and latitudinal gradients. For this purpose, we selected eleven lakes north and south of the Alps as test sites, covering a wide range of morphological and hydrological features, trophic status, and climatic conditions.
We use Sentinel-2 products to derive information of the light penetration and turbidity at a high temporal and spatial resolution. Our workflow is based on remote sensing image processing, field data acquisition, model setup and calibration via data assimilation, and real-time operational model publication on an open-access web-based platform. Sentinel-2 Secchi depth products obtained with state-of-the-art algorithms (e.g., QAA) will be validated with monitoring data. Dedicated field campaigns will be conducted to improve performance by means of generalized inherent optical properties for lakes in the Alpine region. Such products are crucial to constrain and improve the hydro-thermodynamic models, as transparency information is used in the heat fluxes models to parameterize the distribution of the incoming solar radiation in the water column and hence to correctly reproduce the lake thermal structure.
Similarly, existing algorithms for TSM retrieval will be tested and optimized for the use case of Sentinel-2 and the Alpine region. The resulting TSM maps are used to validate the simulated flow field and understand the transport dynamics in the lakes. To this aim, we use a Lagrangian particle tracking module coupled with the three-dimensional hydrodynamic model. Spatial patterns identified in Sentinel-2 images will serve as a proxy for the particle tracking seeding area and particles concentration. This allows tracking the evolution of detected spatial structures in Sentinel-2 image driven by turbulence and mixing processes in the lake. The accuracy of this method will be assessed by comparing the predicted evolution of particles paths with the succeeding Sentinel-2 TSM products.
For dissemination of and user interaction with the combined Sentinel-2 products and hydrodynamic simulations, we will provide hindcasting, real-time, and forecasting functionalities in a web-based platform on the basis of Datalakes (https://www.datalakes-eawag.ch/). This will allow open access to all results, provide a common tool for scientists, decision makers and the broader public, and improve the management of lakes in the Alpine lakes as well as the public perception of environmental processes in their immediate living space.

Water quality is a key worldwide issue relevant to human consumption, food production, industry, nature and recreation. In fact, monitoring and maintaining good water quality are pivotal to fulfilling the UN Sustainable Development Goals and enshrined in European policy through the Water Framework Directive (WFD) and the Marine Strategy Framework Directive (MSFD). Inland, transitional and coastal waters are increasingly threatened by anthropogenic pressures including climate change, land use change, pollution and eutrophication, some of which remote sensing can provide useful and continuous monitoring data and diagnostic tools for.
The European Copernicus programme includes satellite sensors designed to observe water quality and serves data and information to end-users in industry, policy, monitoring agencies and science. Three Copernicus services, namely Copernicus Marine, Copernicus Climate Change and Copernicus Land, provide satellite-based water quality information on phytoplankton, coloured dissolved organic matter, and other bio-optical properties in oceanic, shelf and lake waters. Though the transitional waters are partly covered by CMEMS coastal service, the approaches are distinct in the different services.
Responding to global needs, the H2020 Copernicus Evolution: Research for harmonised and Transitional water Observation (CERTO) project (https://www.certo-project.org/) is undertaking research and development to produce harmonised and consistent water quality data suitable for integration into each of these Copernicus services, and, thus, extend support to the large communities operating in transitional waters such as lagoons, estuaries and large rivers. This integration is facilitated by the development of the CERTO prototype, a Software-as-a-Service (SaaS) that contains modules on improved optical water classification, improved land-sea interface and atmospheric correction algorithms, and a set of selected indicators. The development of suitable indicators that respond to user needs is of utmost importance to demonstrate the added value of the CERTO upstream service to potential users and stakeholders in the downstream service domain. By providing a harmonised capability across the Copernicus services, the CERTO prototype will enable the evaluation of these indicators for the continuum from lakes to deltas and coastal waters and support intermediate and end-users in industry and policy sectors, while ensuring compliance with their own monitoring requirements.
To demonstrate the value of CERTO outputs, six case study areas are selected: i) Danube Delta; ii) Venice Lagoon and North Adriatic Sea; iii) Tagus Estuary; iv) Plymouth Sound; v) Elbe Estuary and German Bight; and vi) Curonian Lagoon. Eighteen local and national stakeholders in the six European countries where the CERTO case study areas are located have been interviewed to identify user needs in terms of the contents and relevance of the CERTO prototype.
Initial analysis of the user requirements collected points to a need for: i) improved products with respect to spatial and temporal resolution, ii) water quality indicators that aggregate data and iii) help in decision making and reporting, such as for the EU WFD and MSFD. To address these needs, several indicators are being developed within CERTO that use satellite-based estimations of water turbidity, suspended particulate matter and chlorophyll-a concentration, and include region-specific mean values, anomalies, percentiles (e.g., chlorophyll-a 90th percentile) and trends. Two indicators are based on turbidity and suspended matter and aim at aiding the planning and management of industry and local authorities: one will allow the analysis of the maximum turbidity zone (or high loads zone), the second one will aim at characterising the dredging events and impacts in the study areas. Another indicator based on the phenological analysis of phytoplankton blooms (i.e., bloom timing) that occur in these transitional regions is also under development. The aim of this indicator is to further understand ecosystem functioning and to provide support for the implementation of additional phytoplankton metrics for the EU WFD. Based on Sentinel-2 and Sentinel-3 data, these indicators are transferable and comparable across time and space and provided in near-real-time to provide faster response. In addition, a more complex indicator is under development, i.e., the Social-Ecological System Vulnerability Index (SESVI), which integrates local knowledge and data, 3rd party modelled and satellite data as well as CERTO outputs, to characterise the main pressures in the case study areas and highlight hotspots of vulnerability in lagoons and estuaries due to human pressure and climate change.
This paper presents the suite of CERTO indicators that aim to better support water resources management and decision-making, and shows the progress achieved thus far.

Chlorophyll-a concentration, as a proxy of phytoplankton biomass, is a key variable to monitor the highly dynamic transitional waters in coastal areas, often subjected to anthropogenic pressures that severely modify its ecological status. Sentinel-3 data is especially suited for this purpose in the large coastal lagoons characteristic of the Mediterranean floodplains. Its spatial, spectral and radiometric resolutions, together with a short revisit time, allow to monitor the spatiotemporal changes of the phytoplankton populations in these ecosystems in response to diffuse pollution, as well as the abrupt changes occurring after extreme meteorological events, such as floods, which alter its hydrodynamics and water composition.
We present the validation of the Sentinel-3 Chlorophyll-a concentration product ([Chl-a]), produced by the C2RCC processors available in the SNAP software, in two Mediterranean coastal lagoons in Eastern Spain: Albufera de Valencia, a shallow hypereutrophic brackish lagoon with an ongoing restoration plan to limit its nutrient content; and Mar Menor, a hypersaline mesotrophic lagoon which is undergoing an accelerated eutrophication, severely affecting its fisheries and its important recreational uses.
For this validation exercise, a set of 1413 and 185 in situ [Chl-a] samples were available in the Mar Menor and Albufera lakes, respectively. In the Mar Menor the in situ data was measured between August 2016 and October 2019, while for the Albufera the time span was between January 2016 and February 2018.
A total of 1142 Sentinel-3/OLCI images, from April 2016 to February 2020 were processed with C2RCC and filtered using the processor’s quality flags. The match-up points were statistically filtered and the correlation with the C2RCC [Chl-a] product was analyzed for the whole dataset, per lake and in the time series.
In the Mar Menor lagoon, with in situ [Chl-a] ranging from ~0.1 to ~25 mg·m-3, the C2RCC accuracy (R2= 0.69; RMSE= 4 mg·m-3) was acceptable for the spatial and temporal monitoring of this variable, closely following the time evolution of [Chl-a] in the studied period and identifying the bloom episodes and abrupt changes after flooding events. On the contrary, in the Albufera lagoon, C2RCC systematically underestimate [Chl-a] to an order of magnitude less than the in situ data, with large retrieval errors (R2=0.40 ; RMSE= 44 mg·m-3), precluding the spatio-temporal monitoring of [Chl-a] and suggesting that the C2RCC could not be appropriate for these eutrophic or hypereutrophic ecosystems.

Phaeocystis globosa is a nuisance haptophyte species that forms annual blooms in the southern North Sea and other coastal waters. At high biomass concentration, these are considered harmful algal blooms due to their deleterious impact on the local ecosystems and economy, and are considered an indicator for eutrophication. In the last two decades, methods have been developed for the optical detection and quantification of these blooms, with potential applications for autonomous in situ or remote observations. However, recent experimental evidence suggests that the interpretation of the optical signal and its exclusive association with P. globosa may not be accurate. Specifically, in the North Sea, blooms of P. globosa are synchronous with those of the diatom Pseudo-nitzschia delicatissima, that are found growing over and inside the P. globosa colonies. P. delicatissima is another toxic harmful bloom-forming species with similar pigmentation and optical signature to P. globosa.

In this study, we combine new and published measurements of pigment composition and inherent optical properties from pure cultures of several algal and cyanobacterial groups, together with environmental spectroscopy data, to identify the pigments generating the optical signals captured by two established algorithms: (1) The classification tree based on the positions of the maxima and minima of the second derivative of the water-leaving reflectance data; and (2) the Chlorophyll c3 (Chl c3) concentration estimation with a reflectance exponential baseline height. We further evaluate the association of those pigments and optical signals with P. globosa.

Our results show that the interpretation of the pigment(s) generating the optical signals captured by both algorithms were incorrect and that the published methods are not specific to P. globosa, even in the context of the phytoplankton assemblage of the southern North Sea. The positions of the maxima and minima in the second derivation of the water-leaving reflectance are defined by the relative concentrations of total Chl c and photoprotective carotenoids (PPC), and not Chl c3 and total carotenoids, as previously suggested. Similarly, the exponential baseline height captures the signal of total Chl c concentration, and cannot isolate the signal from Chl c3 due to the large overlap in the Soret band center position in the Chl c family. Additionally, the position of the minima and maxima of the second derivative can be affected by the presence of Chl b and environmental conditions influencing PPC concentration.

More fundamentally, we found that the optical and pigment signatures of Phaeocystis species are part of a broad pigmentation trend across unrelated taxonomic groups, related to chlorophyll c3 presence. Based on a large database of pigmentation patterns from pure cultures, we observed that the presence and amount of Chl c3 is positively correlated with the concentration of total Chl c and negatively correlated with PPC concentration. This has important consequences for the interpretation of pigment and optical data, particularly in environments where multiple species with similar pigmentation pattern co-occur, as observed in the southern North Sea during P. globosa blooms.

The available information on the relative contribution of cell biomass and pigments to the total pool from diatoms containing the Chl c3 and P. globosa suggests that it is not possible to unequivocally assert that the signal is generated by P. globosa. This is a consequence of year to year variation in the relative cellular biomass of these species during the bloom, the progressive colonization of P. globosa by P. delicatissima along the bloom development, and the low pigment to cellular biomass from P. globosa when compared to the diatoms.

We therefore propose and validated an algorithm to estimate the fraction of Chl c3 from the total Chl c pool, as it carries information on the presence of this pigment and the relative dominance of species presenting the pigmentation pattern of high total Chl c and low PPC. In the southern North Sea, this pigmentation pattern is only observed in P. globosa, P. delicatissima and Rhizosolenia species, the first two being HAB species and dominating the biomass and pigment signal. The Chl c3 fraction in the southern North Sea therefore can be interpreted as an indication of the the relative dominance of HAB species. The new algorithm suffers minimal influence from co-occurring pigments (e.g., Chl b, other forms of Chl c, carotenoids) and can be applied to absorption or reflectance data, with potential for application to the next generation of aquatic space-borne hyperspectral missions. We further elaborate general recommendations for the future development of algorithms for phytoplankton assemblage composition, considering the biology, ecology, optical signal and interpretation.

Abstract: Superficial aquatic environments, including oceans, lakes and rivers, contain a great diversity of particulate and dissolved materials. The water-leaving radiance is directly driven by the optical properties of those in-water materials interacting with light, also known as optically active water constituents (OAWC). In turn, their inherent optical properties (IOPs), such as the absorption coefficient or the scattering matrix, are dependent on the nature of the particles in suspension (i.e., microalgae, sediments). More precisely, the IOPs of suspended sediments depend on their mineralogy including the spectral complex refractive index and size distribution. Nevertheless, the relationship between remotely measurable water reflectance and the IOPs is still to be better elucidated in turbid and very turbid waters. One of the goals of this study was to reassess the IOPs-reflectance forward model over a wide range of water turbidity, and accounting for the polarized nature of light. Moreover, a special focus was paid to evaluate the role of the viewing geometry (sun and viewing angles, and relative azimuth angle between Sun and sensor) and to provide the uncertainty attached to such widely used forward model.
A second part of this work was dedicated to hyperspectral and multispectral analysis of the performances of retrieval algorithms based on the developed forward model. A specific inversion scheme was applied to a series of in situ data sets of moderate to highly turbid waters. Results showed the need to consider the actual multimodal size distribution and spectrally dependent refractive index to accurately reproduce hyperspectral observations. However, the presence of very coarse particles (> 20 µm) produces ambiguities in the retrievals due to their minimal contribution to the water-leaving radiance. Conversely, those findings demonstrate the sensitivity of the measured reflectance to size distribution, thus providing a framework for size distribution retrieval from space. Based on those results, we argue that physically based analysis of the signal remains a fundamental step to gain more genericity and applicability of suspended sediment retrieval algorithms enabling to reconcile the exponentially increasing number of regional algorithms.

The main objective of the H2020 funded project Water Quality Emergency Monitoring Service (wqems.eu/) is to provide operational water quality information to environmental authorities and water utilities industry in relation to the quality of the ‘water we drink’. To reach this goal, the project focuses its activities on monitoring of lakes using a variety of information sources.

The project includes altogether five pilot areas in Greece, Italy, Spain, Germany, and Finland. This work focuses on Lake Pien-Saimaa which is a medium-size lake in southeastern Finland. It is an important source of fresh water for the city of Lappeenranta with water intake located in the southern part of the lake. Lake Pien-Saimaa is fragmented and includes several islands, and it exhibits variable and site-specific water quality features. Lake Pien-Saimaa has substantial intrinsic value to the local population and its many small islands and beaches serve as a location for many holiday houses and recreational activities. The main anthropogenic pollution sources (e.g., phosphorus load) are the surrounding agricultural, and peat production areas and the industrial point-source of Kaukas pulp and paper mill. The main concern in the lake is monitoring algal blooms for early warning purposes.

The EO based data flow utilizes Copernicus Sentinel-2 images processed by the Finnish Environment Institute (SYKE). SYKE provides information on water chlorophyll a concentration and turbidity as maps and timeseries from small areas in various parts of the lake. In situ observations are gathered from bottle samples that are analyzed in laboratory and with instruments installed to an automated water monitoring station located near the water intake. These data are used for the validation and calibration of satellite data. The spatial and temporal behavior of water quality parameters is visualized for end users through TARKKA map service operated by SYKE.

The poster will present results from the EO processing, in situ data collection and the visualization of the results.

This project has received funding from the European Union’s Horizon 2020 Research and Innovation Action programme under Grant Agreement No 101004157.

Satellite images play a crucial role in monitoring Earth’s oceans, especially when it comes to oil spills. Traditionally, detection methods use Synthetic Aperture Radar (SAR) images that allow the detection of oil spills independent of clouds or daylight. However, SAR based methods are limited by wind conditions, as well as, look-alikes. Multispectral satellite images are perfect tools to fill this gap given that they allow the detection of pollution when weak or strong winds do not allow the use of SAR images. For this, a case of oil spill contamination is investigated in an inland lake in northern Greece using Sentinel-2 and PlanetScope multi-spectral images. This case is characterized by a small sample of known oil spills, making this study even more challenging. First, we implement different atmospheric corrections to acquire the remote sensing reflectance for the multispectral bands. Our sensitivity analysis shows that the detection capability for oil spills in not constrained only to the visible (VIS) part of the spectrum, but also extends to the Near Infrared (NIR), as well as, the Short Wavelength Infrared (SWIR). Among these, the NIR (833 nm) and Narrow NIR (865 nm) seem to have the largest sensitivity to fresh water oil spills. Additionally, the oil spills investigated tend to enhance the remote sensing reflectance for the NIR and SWIR part of the spectrum, but reduce it for the VIS bands, with the exception of Red (665 nm), which has a more ambiguous behavior. Given the reduced known oil spill cases (just two) for this study a pixel based machine learning approach is implemented instead of an object based one. Furthermore, the size of the oil spills will determine the choice of the bands, given that low resolution bands tend to reduce the pixel sample, and high resolution bands are limited by their availability (only four are available). Finally, the chosen bands are feed into a Deep Neural Network with two hidden layers for processing and the optimal hyperparameters are investigated. Despite the limited oil spill sample, the results are encouraging, showing a good detection capability.

Hydrological models are a widely used tool to explore which small- and large-scale interventions are suitable to effectively manage water resources, and to gain understanding of this coupled human-natural system. While many processes in the hydrological system can be generalized at large-scale, research in the realm of social hydrology has shown that many important decisions are made at the local level by a highly heterogeneous population, such as reservoir managers and farmers. Effective simulation of these decisions and their effects on the system, thus also involves simulating the coupled human-natural system and its feedback loops simultaneously at a local level and basin scale. Fortunately, an increasing number of high-resolution datasets have become available, for a large part driven by satellite observations, facilitating simulations at high resolution. Examples of datasets and methods include delineation of functional crop fields with machine learning using data from Sentinel, WorldView-3, and high-resolution SAR, associated field-scale availability of cropping and irrigation patterns, as well as high-resolution soil moisture data from downscaling of passive microwave observations and SAR.

Therefore, to capitalize on these advances, CWatM has been developed in several ways. First, we have enabled the ability of CWatM to be run at 30’’ resolution (< 1km at the equator), with examples in the Bhima basin (India), Burgenland (Austria), as well as in China and Israel. Associated developments include specific crops and fallowed land, calibrated reservoir operations, water distribution areas from reservoirs (command areas) or rivers (lift areas), canal leakage, as well as explicit source- and sector-specific water demands. An updated calibration scheme calibrates subbasins in a cascading fashion from upstream to downstream and generates parameter maps for each subbasin. The calibration scheme uses and evolutionary computation framework in Python (DEAP package) and the modified version of the Kling-Gupta Efficiency as objective function for comparing simulated with observed streamflow at subbasin scale.

Furthermore, we made advances in increasing the computational speed of CWatM. For example, CWatM now uses MODFLOW 6 through its Basic Model Interface (BMI) interfaced through Python (FloPy and xmipy packages). Using this method, we tested the high-resolution simulation of groundwater at 250m in the Bhima Basin (India) and 100m in Burgenland (Austria). Here, we use MODFLOW to represent physically one aquifer layer and to simulate groundwater interactions with soil and surface water bodies, as well as pumping demand at a daily timestep. In both areas, the model successfully reproduced observed water table better than at low resolution. In addition, many grid-based calculations, such as the soil-water balance can now be run in parallel on the GPU, enabling a tens of times faster resolution of the soil-water balance, depending on the hardware configuration.

Finally, IIASA developed, in collaboration with IVM-VU, an agent-based model (ABM) that simulates millions of individual farmers and their bi-directional interactions with CWatM at field scale, parameterized with the aforementioned high-resolution satellite products. However, because each agent requires their individually operated soil-water balance, a very high-resolution hydrological grid is required, limiting the ability of CWatM to be run in large basins. Therefore, to effectively manage this, we introduced land management units in CWatM. In this concept, CWatM is still run at the 30’’ resolution with 6 different land use types, but crop land use types are further subdivided based on land ownership types and become dynamically sized hydrological response units (HRUs) within the grid cell. These land management units can be independently operated by farmers through the ABM. In this manner, all land management practices (e.g., crop planting date and irrigation) and soil processes (e.g., percolation, capillary rise, and evaporation) are simulated independently per farmer, thus allowing simulation of multiple independently operated farms within a single grid cell. Runoff and percolation to groundwater are aggregated from all HRUs within a grid cell to simulate groundwater and river discharge at the grid scale. This enables CWatM to simulate the bi-directional interaction of individual farmers with the hydrological system and their adaptive behaviour at the true farm scale, while still allowing simulation of the hydrological process at basin scale. We show an example of ~11.1 million farming households in the Krishna basin in Indian, simulated on a personal laptop. Calibration with streamflow shows good model performance, and as a next step, we plan to further calibrate with high-resolution soil moisture products at ~100m resolution.

Cyanobacteria are a persistent problem in inland waters. They hamper the use of water for recreation and drinking purposes. Being odorous and foul-looking they are also an unwanted guest in urban waters. Water Insight has supported a number of water managers in the Netherlands and abroad to monitor the onset of cyanobacteria blooms and take early and appropriate action.
Based on simultaneous measurements of cyano-chlorophyll (with lab fluoroprobe) and optical Phycocyanin (WISPstation) we were able to establish a good relationship between the two parameters. This relationship could be extended to biovolume based on a large Dutch database of measurements. Water managers often prefer biovolume as indicator for the abundance of cyanobacteria. The general validity of the conversions should be investigated further.

We present 3 use cases explaining the usability of our concept to monitor the blooms with the most suitable combination satellite observations and our proprietary optical sensors.

Case 1: Bathing water monitoring
A WISPstation was used in the two small lakes (“Agnietenplas” and “Bosplas”) in the Netherlands to demonstrate the added value of continuous in-situ monitoring of the growth and decline of cyanobacteria. The optical data record clearly shows the added value of high frequency measurements opposite to a 2-weekly sampling frequency. Short-term peaks are recognised and the bathing water can be opened or closed on a daily basis instead of a 2 weekly basis.

Case 2: Determination of the representativeness of WFD monitoring stations
Lake Lauwersmeer suffers from high-concentration blooms. The purpose of the Water Framework Directive is to take measures to improve the water quality to a ‘good’ ecological status, however, sparse sampling and therefore less insight makes it difficult to take effective measures. In a pilot of H2020 e-Shape, satellite data was used to map the EO-base phytoplankton biomass for WFD reporting in Lake Lauwersmeer, and study the representativeness of the existing monitoring stations.

Case 3: Early warming for nuisance of blooms in a recreational harbour.
In this case in-situ optical measurements of the WISPstation serve two purposes: the high-frequent measurements are used as early warning of upcoming blooms, while the spectral data serve as calibration of atmospheric correction in this turbid and cyanobacteria infested lake Volkerak. Using this technique significantly improved the quality of Sentinel-2 BOA reflectances. Based on the early warning for blooms, the water manager temporary closes a small harbour preventing the nuisance blooms to enter.

EOMORES has received funding from the European Union’s Horizon 2020 research and innovation programme grant agreement 730066
e-Shape has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement 820852

The European Union Water Framework Directive (and similar directives) requires countries to monitor and report on the ecological status of inland and coastal water bodies, through biological, chemical and physical indicators. Countries reporting on a large number of water bodies struggle to collect sufficient observations to represent seasonal and interannual variability, particularly in dynamic systems influenced by terrestrial runoff. Monitoring of certain indicators can be complemented using Earth observation to fill such observation gaps and inform better management. Satellite remote sensing is particularly complementary and can be achieved at relatively low cost, but water quality products from satellites are still limited to relatively wide, open water bodies.

In shallow coastal waters and intertidal zones, runoff from farming nearby land, water treatment works outflow and untreated sewage can lead to nutrient conditions in which macroalgae flourish and out-compete other beneficial plant life such as Zostera seagrasses. Such shifts can disturb local ecology and lead to loss of biodiversity, reduce important carbon sequestration and negatively affect the blue economy.

With the use of high resolution EO data such as Sentinel-2, together with image processing and machine learning techniques, we are able to observe the areal coverage of vegetation within a tidal lake in the southwest UK and estimate the seasonal variation in cover from a 5-year time series of Sentinel-2. Photography taken from unmanned aerial vehicles additionally provides a very high resolution (~4 cm) view for evaluation or creation of training data, and an estimate of macroalgae coverage itself. Quadrat surveys (which are the accepted reference method) in the region provide further information, but at limited spatial and temporal coverage. The differences in these three levels of observation (satellite, UAV, quadrat) are discussed with suggestions on how they might be reconciled in future so that the wide area and regular temporal coverage that satellites offer can be used in reporting.

Using a clustering approach with the Sentinel-2 data we were able to assign pixels within the lake to classes relating to mud, water and vegetation. Aggregating into seasonal periods suggests that the vegetation coverage within the lake ranges from approximately 5 % of the intertidal area in winter up to 60 % in summer. UAV data, which only covers a portion of the lake, suggests much lower summer coverage of 8 %, whilst the quadrat reference method reports 68 %. To be able to use EO techniques in future WFD activities these methods will need calibrating and agreement with relevant bodies so that the spatial and temporal benefits of EO data can be fully utilised.

Lakes perform a multitude of functions, from regulating water flow and quality to providing food and income from fishing and tourism. Lakes moderate the local climate and provide water for drinking and irrigation. All of these functions are being affected by human actions. Pollution by untreated waste water and fertilizer use cause eutrophication and disturb the ecosystem’s balance. Warming of the climate causes enhanced evaporation, changing precipitation patterns, and affects the stable layering of lakes, which, in turn, affects the ecosystem by changing the availability of oxygen and nutrients. And these effects influence each other in various and complicated ways. Thus the warming climate may exacerbate the influence of increased nutrient influx.
People living near lakes are directly affected by these changes, some of which can be observed using satellite instruments. Monitoring the quality of lake water can help understand processes leading to changes in lakes, which will aid the development of mitigation or adaptation strategies. Moreover, finding relationships between satellite data and disease or other risks allows their prediction, providing the basis for an early response.
Here, we present first results on the monitoring of the greenness of lakes for three different applications, each addressing an aspect of human health. First: Blue algae, or cyanobacteria, thrive in nutrient-rich, warm waters. In high numbers, they outcompete other algae and plants and have toxic effects on animals and humans. Second: The water hyacinth, which has evolved into a major disturbance to water traffic, fishery, and lake ecosystems within a few decades. Dense water hyacinth mats provide breeding grounds for vectors of malaria and leishmaniasis. Third: Phytoplankton, whose abundance was shown to be related to cholera incidences in various regions in Asia and Africa, as the bacterium responsible for the disease associates with phytoplankton.

Monitoring water turbidity in water bodies provides useful information on hydrological processes occurring at the watershed scale as well as on the state of aquatic ecosystems including bacteriological contamination. Quantification of suspended sediment is also important for reservoirs management since it allows to monitor silting that can affect dams functioning while providing important information for water treatment. Remote sensing provides a useful tool for monitoring inland water at the regional scale but only recent satellites provide the spatial and temporal resolution necessary to follow the dynamics of small water bodies.

This study is focused on the Sahelian region where ponds, lakes and reservoirs play a major role for populations. Given their small size, their important temporal variability, and the scarcity of in-situ monitoring network, their dynamics and water quality information is not available at the regional scale. In addition, Sahelian water bodies are very reactive to climate and human forcing and display complex and sometimes unexpected behaviours, like increasing trends in water area across the Sahel, which question their future evolution in a context of environmental changes and demographic increase.
We explore the capability of Sentinel2 optical sensor MSI to retrieve information on waterbodies variability at the large scale, using Google Earth Engine to processes several Sentinel2 tiles. Overall, 1672 Sahelian lakes are analysed and compared to other 5666 lakes in semi-arid regions worldwide.

Water reflectance in the visible and NIR bands significantly vary across different lakes and can reach extremely high values (higher than 0.4 in the NIR band) in some lakes, as for example for several lakes located in Niger which are among the brightest in the world.
In-situ measurements over some of these lakes highlights the high concentration of suspended particulate matter (SPM) which increases water reflectance. In addition, SPM are mainly composed of fine kaolinites , that display low absorption coefficient and so high reflectance. Finally the important fraction of very fine mineral particles (a major volumetric mode is found at 200-300 nanometers) may induce an increased diffusion and a higher back-scattering, which both contribute to increase reflectance. High aerosols conditions and sunglint effects are efficiently masked by the image processing and the post-processing applied (based on thresholds on the MNDWI index and on the reflectance in the blue band) and do not significantly affect the results reported.

At the regional scale the brightest lakes are identified as relatively small lakes situated in area with low vegetation cover, where erosion and sediment transport is more likely. However, the contrary isn’t not always true and low reflectance values can be encountered on small lakes in low vegetated areas, as it happens, for example, for lakes fed by water table or by the flooding of the Niger river.
Observations by high resolution remote sensors such as Sentinel2 are thus an efficient tool to derive information on water color spatial variability in relation to eco-hydrological characteristics at the regional scale.

Lake water quality is a key factor for human wellbeing and environmental health being affected by climate change, anthropogenic activities such as urban and domestic wastewater charges into in inflowing streams, as well as agricultural activities. Generally, in situ measurements are traditionally conducted and widely accepted as instruments for water quality monitoring. However, in many regions, classical monitoring capacities are limited and in case of large water bodies they lack monitoring at the required spatial and temporal scales. In this context remote sensing has great potential and can be used for assessing spatio-temporal dynamics of water quality in a cost- effective and informative manner.
Copernicus Sentinel-3 OLCI instrument launched in February 2016 covering 21 spectral bands (400-1200nm). It has been providing accessible products since 2017, which enables monitoring water bodies at 300m resolution and almost daily scales.
Lake Sevan (40°23‘N, 45°21‘E) is located in Gegharkunik province in Armenia at an altitude of 1900 m a. s. l.. Lake Sevan is Armenia’s largest water body and the largest freshwater resource for the whole Caucasian region. At present, lake water quality is progressively deteriorating due to eutrophication, water level fluctuations, and climate warming and suffers from massive cyanobacterial blooms. Hence, it is very important to study key water-quality variables (e.g. Chl-a, turbidity, harmful algae bloom etc.) describing the ecological status of the lake, and their annual and seasonal dynamics. We emphasize that remote sensing can significantly contribute to this monitoring demand.
In situ measurements having been implemented since 2018 in the frame of German-Armenian joint projects (SEVAMOD and SEVAMOD2) providing detailed information on water quality of the lake on a local scale. These data are, on the one hand, well suited for a basic characterization of the problem but do not provide enough data for tracking the qualitative changes of the water in space and time.
Hence, our research aimed at assessing the seasonal and spatial water quality dynamics in Lake Sevan over 5 years covering 2017-2021 by using Copernicus Sentinel 3 products.
The satellite data processing engine for water quality assessment eoLytics was used for Sentinel 3 data processing. EoLytics is based upon the MIP Inversion and Processing System, that was initially developed at DLR from 1996 and has been continued since 2006 by EOMAP. The fully physics based sensor-generic algorithm in MIP do not require in-situ data for calibration.
The full range of daily incoming time series for Lake Sevan that is free of clouds includes 474 scenes. Using EoLytics algorithms the data were processed for the water quality parameters: Chl-a, total suspended matter (TSM) and harmful algae bloom (HAB-Indicator). It is noteworthy that Chl-a and TSM processing algorithms provide fully quantitative outputs while the HAB algorithm provides a semi-quantitative indicator.
Field campaigns of in situ measurements have been conducted since 2018 on a monthly basis. andIn situ measurements are available at two locations for 2018-2020 years and at three locations for 2021.
This study envisaged the following steps: (i) the analysis of the annul and seasonal remotely sensed data in order to reveal the spatial-temporal characteristics of water quality; (ii) the comparison of in situ measured and remotely retrieved data via regression analysis in order to understand the relationship of the data received from the different sources.
The spatial-temporal analysis of the seasonal characteristics of Chl-a follows typical plankton succession dynamics in large lakes and usually shows maximum values in summer (June and July). However, the seasonal dynamics of Chl-a differ between years and are most likely driven by meteorological dynamics. This link between meteorological variables and plankton dynamics is a key aspect for climate impact assessments and are hardly visible in traditional monitoring data but well observable in remotely sensed data due to its higher temporal resolution. The spatial patterns in the lake point to a large influence from external nutrient inputs as high Chl-a values are repeatedly observed in the vicinity of polluted inflows.
The HAB repeats the overall seasonal trend we have for chlorophyll-a with the highest appearance in 2018, then a bit less in 2019. It follows with much more low appearance in 2020 and 2021. 2017 gives the lowest picture of the HAB occurrence.
An initial comparison of in situ measured and remotely sensed Chlorophyl-a via a linear regression revealed a significant relationship with a relatively low error (RMSE=0.403).
At this stage it can be concluded that the maximum of the Chl-a content during these 5-year period slightly shifted from August to June and was regularly associated with the formation of HABs. The validation of the remote sensing data needs ongoing efforts in order to facilitate a deeper analysis of seasonal trends and spatial-temporal patterns. The observations based on Sentinel-3 sensors provide extremely valuable information on water quality dynamics in Lake Sevan and complement the results from traditional monitoring. The long-term monitoring strategy should therefore exploit the strengths of both approaches and remote sensing is considered to be a key aspect of the foreseen monitoring program for this highly sensitive and important water body.
Acknowledgments
This study was supported by following projects:
1. SevaMod - Project ID 01DK17022 - Funding Institution: Federal Ministry for Education and Research of Germany "Development of a model for Lake Sevan for the improvement of the understanding of its ecology and as instrument for the sustainable management and use of its natural recourses"
2. “SevaMod2: Project ID: 01DK20038 - Funding Institutions: Federal Ministry for Education and Research of Germany and 91.5% of planned costs), Ministry of Environment of the Republic of Armenia (8.5% of planned costs) “Building up science-based management instruments for Lake Sevan, Armenia”
3. Project ID - 20TTCG-1F002 - Science Committee of the Ministry of Education and Science, Culture and Sport of RA "The rising problem of blooming cyanobacteria in Lake Sevan: identifying mechanisms, drivers, and new tools for lake monitoring and management"
4. Project ID - 21T-1E252 - Science Committee of the Ministry of Education and Science, Culture and Sport of RA "Assessing spatio-temporal changes of the water quality of mountainous lakes using remote sensing data processing technologies".

The Water Framework Directive (2000/60/EC) (WFD) states that all European Union (EU) members must implement the monitoring and estimation of the ecological status of their territorial inland water bodies. It demands to classify the status in 5 classes from “very bad” to “high” and aims to achieve at least “good” ecological status of inland waters by 2027 by all required means. However, monitoring of the several ecological parameters required by the WFD are based on in situ data sampling and further laboratory analysis, that are both time- and money-consuming. Hence, this cannot be achieved in timely and frequent manner at a country-scale. Therefore, satellite observation of water quality appears to be a promising and efficient tool to achieve the WFD requirements (Papathanasopoulou et al., 2019). Nonetheless, there is still a need to improve the accuracy of satellite-derived products used to classify the water bodies status, especially for each water quality parameters that can be remotely sensed: chlorophyll-a concentrations ([chlo-a]) values, Secchi-disk depth, turbidity, suspended matter concentrations. (Giardino et al., 2019).

In order to evaluate the relevancy of the current satellite products for ecological status monitoring, this study was based on numerous French lakes sites where field data were previously collected. A dataset was composed with in situ measurements from the french WFD regulatory monitoring network and the long-term Observatory on Lakes (OLA), as well as from other public institutes (research or territorial management). This dataset concerns ~325 sites from 2014 to 2017. Over the period covered by Lansat8 and Sentinel 2, this dataset includes ~1000 to 1900 chlorophyll-a concentrations ([chlo-a]) values, Secchi-disk depth, turbidity, suspended matter concentrations. Corresponding satellite data products were generated from Sentinel-2 and Landsat-8 imageries through our processing chain: first, level 2A water reflectances were produced with the atmospheric correction (AC) algorithm “Glint Removal for Sentinel-2 like data” (GRS) (Harmel et al., 2018), second, water reflectances images were masked based on watermask and cloudmask computed with sentinel-hub’s “s2clouless” for S2/MSI, and the original cloudmasks for Landsat 8. To evaluate the quality and representativeness of the satellite products, matchup comparisons were performed.

Our results demonstrate that, in certain environments and circumstances, sunglint signal can represent a major part of the water leaving reflectances. It can lead to a 10-fold bias on water quality estimations, and validate the importance of including a sunglint correction step, even though no consensus exists in the choice of a particular atmospheric correction algorithm (Pahlevan et al., 2021). Focusing on [chlo-a] retrieval, we implemented several widely used algorithms from literature, and adapted them to Landsat 8 and Sentinel 2 data. We calculated [chlo-a] following several modalities: (i) with the original paper’s calibration, (ii) after applying calibration to the region of interest, and (iii) with calibrations defined for each OWT by (Neil et al., 2019). We also implemented a spectral angle mapper method (SAM) to identify Optical Water Types (OWT hereafter) as defined by (Spyrakos et al., 2018), like it was recently implemented for MERIS by (Liu et al., 2021).
For altitude lakes in the Alps mountain chain, classified as clear oligotrophic waters, ocean-colour algorithm OC3 performs best, detecting low [chlo-a] in the range of 1 to 10 µg/l (MAE < 2.6 µg/L, RMSE < 4.8 µg/L, MAPE < 65 %, SSPB(signed bias) < 15%), which is comparable to recent neural-network algorithms. In meso- to eutrophic lakes, several algorithms performed satisfactorily such as red-fluorescence-, NDCI- or 2-3 bands- based algorithms, but with variable accuracies depending on sites. In a few lakes of Brittany and Aquitaine regions, optically classified as eutrophic to hypertrophic turbid lakes, performances are good enough to distinguish periods of blooms, and the shifts between ecological status from “moderate”, “poor” up to “bad”. Ranges of water quality parameters associated with the OWT classes as defined in (Spyrakos et al., 2018) are also in good agreement with the in situ data observed on our sites. Moreover, a matching score derived from SAM and OWT classes was implemented to measure similarity with OWT shapes. Matching scores will soon guide the choice of the best-suited algorithm. This matching score was also shown to be complementary to cloud and water masks and provide further ability for masking out pixels that are likely impacted by contributions from upwelling light of the bottom, adjacency effect from the shores, or badly-masked clouds.

This work, whilst ongoing, showed that spectral identification performs well with high resolution satellite data and is useful to optimize algorithm selection. Reasoning by analogy on optical types, we expect to successfully use OWT classification to retrieve [chlo-a] and other parameters on lakes where in situ data is not available. The perspective of this study is to proceed with a census of the ecological states of the French lakes. This can be foreseen as a crucial step to help respect the WFD engagements since field data is scarce or even absent for many sites included in the WFD.

References

Giardino, C., Brando, V.E., Gege, P., Pinnel, N., Hochberg, E., Knaeps, E., Reusen, I., Doerffer, R., Bresciani, M., Braga, F., Foerster, S., Champollion, N., Dekker, A., 2019. Imaging Spectrometry of Inland and Coastal Waters: State of the Art, Achievements and Perspectives. Surv Geophys 40, 401–429.
Harmel, T., Chami, M., Tormos, T., Reynaud, N., Danis, P.-A., 2018. Sunglint correction of the Multi-Spectral Instrument (MSI)-SENTINEL-2 imagery over inland and sea waters from SWIR bands. Remote Sensing of Environment 204, 308–321.
Liu, X., Steele, C., Simis, S., Warren, M., Tyler, A., Spyrakos, E., Selmes, N., Hunter, P., 2021. Retrieval of Chlorophyll-a concentration and associated product uncertainty in optically diverse lakes and reservoirs. Remote Sensing of Environment 267, 112710.
Neil, C., Spyrakos, E., Hunter, P.D., Tyler, A.N., 2019. A global approach for chlorophyll-a retrieval across optically complex inland waters based on optical water types. Remote Sensing of Environment 229, 159–178.
Pahlevan, N., Mangin, A., Balasubramanian, S.V., Smith, B., Alikas, K., Arai, K., Barbosa, C., Bélanger, S., Binding, C., Bresciani, M., Giardino, C., Gurlin, D., Fan, Y., Harmel, T., Hunter, P., Ishikaza, J., Kratzer, S., Lehmann, M.K., Ligi, M., Ma, R., Martin-Lauzer, F.-R., Olmanson, L., Oppelt, N., Pan, Y., Peters, S., Reynaud, N., Sander de Carvalho, L.A., Simis, S., Spyrakos, E., Steinmetz, F., Stelzer, K., Sterckx, S., Tormos, T., Tyler, A., Vanhellemont, Q., Warren, M., 2021. ACIX-Aqua: A global assessment of atmospheric correction methods for Landsat-8 and Sentinel-2 over lakes, rivers, and coastal waters. Remote Sensing of Environment 258, 112366.
Papathanasopoulou, E., Simis, S., Alikas, K., Ansper, A., Anttila, S., Attila, J., Barillé, A.-L., Barillé, L., Brando, V., Bresciani, M., Bučas, M., Gernez, P., Giardino, C., Harin, N., Hommersom, A., Kangro, K., Kauppila, P., Koponen, S., Laanen, M., Neil, C., Papadakis, D., Peters, S., Poikane, S., Poser, K., Pires, M.D., Riddick, C., Spyrakos, E., Tyler, A., Vaičiūtė, D., Warren, M., Zoffoli, M.L., 2019. Satellite-assisted monitoring of water quality to support the implementation of the Water Framework Directive, White paper. Eomores.
Spyrakos, E., O’Donnell, R., Hunter, P.D., Miller, C., Scott, M., Simis, S.G.H., Neil, C., Barbosa, C.C.F., Binding, C.E., Bradt, S., Bresciani, M., Dall’Olmo, G., Giardino, C., Gitelson, A.A., Kutser, T., Li, L., Matsushita, B., Martinez-Vicente, V., Matthews, M.W., Ogashawara, I., Ruiz-Verdú, A., Schalles, J.F., Tebbs, E., Zhang, Y., Tyler, A.N., 2018. Optical types of inland and coastal waters: Optical types of inland and coastal waters. Limnol. Oceanogr. 63, 846–870.

Worldwide freshwater systems are impacted by climate warming and anthropogenic forcing, influencing water level and runoff regimes via changes in precipitation and landuse patterns. Especially for river-connected lake systems these rapid changes might have far reaching consequences where inland nutrient loading might accumulate along the river system and finally lead to destabilization of distant ecosystems like estuaries. Thereby lakes, through their influence on flow regime, might play a critical role in how much and how far local eutrophication events will be transported along the river network. Currently, studies on river-connected lake systems are scarce and largely based on data with both low temporal and spatial resolution. Furthermore existing meta-ecosystem theory rarely takes lake-to-lake connectivity into account. In this study, we modeled how local nutrient input influences phytoplankton and how both propagate along strong or weak connected lakes. These theoretical investigations were accompanied by an extensive field study on lakes located along the Upper Havel-river system in Northern Germany including shallow and deep lakes and covering various flow regimes. We investigated effects of local nutrient loading on regional-scale plankton development along river-connected lake chains. To achieve high temporal and spatial resolution, we measured water constituents combining automated in-situ probes with ground-based, space- and airborne reflectance measurements. The field data show that upstream nutrient input drove phytoplankton development along the entire lake chain due to tight hydrological linkage. Our results suggest that similar point sources can result in profoundly different maximum intensity, spatial range and regional-scale magnitude of eutrophication impacts in lake chains dependent on flow regime and lake characteristics. We highlight the potential of combining in-situ measurements with remote sensing to improve lake meta-ecosystems monitoring.

Inherent Optical Properties (IOPs), such as absorption and scattering, link the biogeochemical composition of water and the Apparent Optical Properties (AOPs) obtained from satellites, including remote sensing reflectance (Rrs). The so-called optical closure analysis between radiometrically-measured AOPs and simulated AOPs from measured IOPs and light-field boundary conditions is crucial for assessing and, ideally, minimizing the uncertainties associated with AOP-to-IOP inversion algorithms. However, this step is complicated due to several factors, e.g., the unknown bias and random errors in the individual measurements, limitations in the sampling of the Volume Scattering Function (VSF) and fluorescence emission, and uncontrolled environmental effects, causing uncertainties in the AOPs and water constituents retrieval.
In this study, we used in-water bio-optical data acquired by an autonomous profiler (WetLabs Thetis) several times a day, as well as Sentinel-3 OLCI radiance and reflectance products to quantify, characterize, and mitigate the uncertainty of Rrs estimates. Various bio-optical sensors, as well as a Conductivity-Temperature-Depth (CTD) probe, are mounted on this profiler. Hyperspectral downwelling irradiance and upwelling radiance (Satlantic HOCR; 189 channels between 300-1200 nm), hyperspectral absorption and attenuation (AC-S; 81 channels between 400-730 nm), backscattering at 440, 532, 630 nm at 117° (ECO Triplet BB3W), as well as backscattering at 700 nm at 117° and Chlorophyll-a fluorescence (ECO Triplet BBFL2w) measured at an offshore research platform in Lake Geneva (Switzerland/France), called LéXPLORE (https://lexplore.info/), were used to address the scientific objectives. The in situ dataset includes 294 high vertical resolution daily profiles for the period between 10/2018 and 5/2020. The quasi-concurrent Sentinel-3 data (within ±2 hr of the in situ measurements) were used to assess the performance of the proposed uncertainty characterization and mitigation. The POLYMER atmospheric correction was used to obtain Rrs. We tested two bio-optical models available in POLYMER: (i) the globally optimized model by Garver, Siegel and Maritorena (GSM01), and (ii) the model proposed by Park and Ruddick (PR05). 41 and 31 matchups are available for the GSM01 and PR05 models, respectively.
The Hydrolight (HL) radiative transfer model was employed to obtain Rrs from the measured IOP profiles. We used a combination of different metrics based on the residuals of IOP-derived and radiometrically-measured Rrs to quantify and characterize the optical closure. Using the raw IOP profiles, our closure study indicated 33% of profiles with both error and bias of < 15% (i.e., good closure), and 18% with an error or bias of > 30% (i.e., poor closure). We then investigated the effect of scattering corrections for AC-S measurements, which only slightly improved the results (38% good and 21% poor closure). Next, we evaluated a simple single-step backscattering ratio (Bp) optimization method based on Rrs residuals, which significantly improved the Rrs optical closure (99% good, and 0% poor closure). The resulting optimized Bp shows a plausible seasonal variation ranging from ~0.005 during winter to ~0.024 during the end of spring and the beginning of summer. Our study confirmed that Bp, or more generally the VSF, is the most sensitive parameter in estimating AOPs from IOPs.
We further investigated the effects of uncertainty characterization (i.e., profile clustering) and uncertainty mitigation (e.g., IOPs correction) on the in situ-derived and Sentinel-3 Rrs matchup analysis. The latter showed a similar pattern to pure in situ analyses, i.e., slight enhancement using AC-S scatter-corrected profiles, and recognizable improvement implementing backscattering optimization. To avoid any overfitting by the backscattering optimization, the AC-S scatter-corrected profiles were used for investigating the effect of profiles clustering on matchup analysis. The results revealed that the uncertainty clustering based on the in situ profiler optical closure exercise can be used for Sentinel-3 matchup analysis, i.e., profiles with good closure indicated better performances based on different metrics as compared with poor closure. The satellite-derived Rrs using both PR05 and GSM01 models showed similar patterns in analyzing the effect of uncertainty characterization and mitigation with only slightly better results employing PR05.
Ultimately, we used profiles with good optical closure in other wavelengths to estimate phytoplankton fluorescence quantum yield in the emission region (670-700 nm). By relating these estimates to irradiance and pigment concentration, we managed to derive realistic diurnal estimates of non-photochemical quenching (NPQ) across the euphotic layer. In doing so, we can explain the limitations of fluorescence-based Chlorophyll-a retrieval algorithms for oligo- to mesotrophic lakes, and characterize the impact of photoinhibition on daily integrated primary production estimates.
Our results, in general, highlight the potential of using autonomous optical profiling as an alternative for automated ground-truthing of AOPs, with the added value of simultaneous IOP measurements. Further research is needed to investigate if improved VSF measurements and hence better estimates of Bp, or the consideration of full polarization in radiative transfer simulations enable improved conclusions from optical closure assessments.

Remote sensing can provide valuable information for monitoring the ecological status of inland waters. However, due to the optical complexity of lakes and rivers, quantifying water quality parameters is challenging. One approach is to use remotely-sensed reflectance to classify inland waters into discrete classes – or Optical Water Types - that correspond to different ecological states. These optical classes can then be used either to inform the selection of the most appropriate water quality retrieval algorithms or as valuable ecological indicators in their own right.

This review aimed to understand how remote sensing has been used to classify the ecological status of inland waters and which classification approaches are most effective, as well as identifying research gaps and future research opportunities. Using a systematic mapping methodology, a search of three large literature databases was conducted. The search identified an initial 174 articles, published between January 1976 and July 2021, which was reduced to 64 after screening for relevance.

Very few papers were published before 2008 but since then publications increased substantially. The number of waterbodies included in the studies ranged from one to more than 1000, with the vast majority of studies including five or fewer waterbodies. There was a geographical bias towards Europe, the US and China, with poor representation across Africa and the rest of Asia. The source of spectral data used for training the classifications was overwhelmingly from satellite data or in situ measurements, with relatively few using data from aircraft or UAVs. The most common satellite sensors used were the Landsat series, MERIS, MODIS, Sentinel-2 MSI and Sentinel-3 OLCI.

The classification frameworks used were primarily based on Optical Water Types or Trophic State Index, but many studies adopted their own bespoke classification schemes. The number of classes varied from 2 to 21, peaking at 3 classes. A variety of classification algorithms were utilised including unsupervised clustering, supervised (parametric and machine learning) methods, and thresholding of spectral indices. Most studies related the optical classes to in situ water parameters, particularly Chlorophyll-a, Total Suspended Solids and Coloured Dissolved Organic Matter. A variety of pre-processing steps were applied prior to classification including normalisation of spectral data and dimensionality reduction techniques such as Principal Component Analysis.

In this presentation, we summarise the strengths and limitations of different sensors, pre-processing methods and classification algorithms for optical classification of inland waters. Our results highlight important gaps, such as the geographical bias in studies and training data. We emphasize the need for greater transparency and sensitivity analysis to understand how decisions made about the choice of sensor, classification algorithm and pre-processing steps influence to resulting optical classes. Recommendations for future research are presented, including the need for standardized approaches to support transferability of methods and scaling up from local to global scales.

Freshwaters play a significant role in the global carbon cycle by degassing large carbon fluxes. It is established that most of this carbon emitted to the atmosphere comes from organic matter degradation during transport and storage in rivers and lakes. This is particularly true for freshwaters in tropical context such as Petit-Saut reservoir (365 km²) in French Guiana, with huge inputs of terrestrial organic matter (litter and drowned forest), high temperatures and humidity (both being aggravating factors of the degradation).
Knowledge about spatial distribution and temporal evolution of dissolved (and particulate) organic carbon (resp. DOC and POC) in this reservoir and its tributaries is fundamental for a better understanding of degassing mechanisms and estimation of GHG emissions. Hence, we tested the potentialities of high spatial resolution multispectral satellite imagery (Sentinel-2 and Landsat 8) for monitoring DOC concentrations in these absorbing tropical waters, using the absorption coefficient of the coloured dissolved organic matter (aCDOM) as a proxy.
Optical properties (aCDOM and above water remote sensing reflectance (Rrs)) as well as water quality measurements (DOC, POC, total suspended matter, chlorophyll-a, etc) were carried out at 25 stations evenly distributed over the entire lake. CDOM absorption was the highest at the mouth of the main tributary (Sinnamary river) and the lowest in the pelagic area, near the dam.
Simulated satellite spectra were computed by convoluting in situ hyperspectral data with the spectral response function of the given satellite sensor (Sentinel-2/MSI or Landsat 8/OLI), and have been compared to atmospherically corrected satellite data. We used several atmospheric correction algorithms (ACOLITE, C2RCC, C2X, C2X-COMPLEX, GRS, iCOR, LaSRC, Sen2Cor) and resulted spectra were highly heterogeneous (depending on the method used), and poorly correlated with in situ spectra. We explain these limited performances by environmental factors, such as the presence of absorbing aerosols (e.g., N2O) or strong adjacency effects (IOCCG, 2018) that are sill hardly resolved by atmospheric correction methods. According to the ACIX-AQUA exercise (Pahlevan et al., 2021), it is indeed not uncommon that atmospheric correction processors failed to retrieve realistic water reflectance in very absorbing waters surrounded by dense vegetation – typically the case of Petit-Saut reservoir, located within the Amazon rainforest.
We tested several semi-empirical and semi-analytical algorithms from the literature to estimate aCDOM at 440 nm ( aCDOM(440) ) from multispectral data. We also designed an empirical algorithm based on Sentinel-2 bands B3 to B5, performances of atmospheric correction processors being reasonable on this part of the spectrum. Even though the retrieval of aCDOM in the absorbing black waters of Petit-Saut remains challenging, most of these recalibrated algorithms seem to be robust enough to variable concentrations of total suspended matter, and provide satisfactory results over the entire range of aCDOM observed during our campaign.
In order to retrieve DOC concentration from remote sensing data, a linear relationship between aCDOM(440) and DOC has been defined and suggest that aCDOM(440) can be used as an efficient tracer to estimate DOC in most of Petit-Saut waters. However, points located in main tributaries or their transition zones do not follow the same relationship, which is known to be water body or river specific (Valerio et al., 2018).
To summarise, we were able to estimate DOC in these tropical black waters from simulated satellite spectra, but there are still several challenging issues to overcome before being able to do it from space – atmospheric correction being the first order source of uncertainty associated with aCDOM estimation in such highly absorbing environments. New measurement campaigns will be conducted i) to enrich our dataset and precise the optical properties of tributaries and transition zones, ii) to study possible seasonal or inter-annual variation of the aCDOM(440)–DOC relationship (Del Castillo, 2005) and iii) to better constrain atmospheric corrections by having in situ AOT measurements.
When the above-mentioned limitations will be overcome, our next objective will be to produce time series of DOC concentrations from Sentinel-2 and Landsat 8 archives. It will help to characterize the spatial and temporal distribution of organic carbon in the area, which is useful to better comprehend organic matter degradation processes and dynamics. Ultimately, this will benefit both public authorities and Électricité de France (the dam manager) in their management of the dam and its reservoir.


References

Del Castillo, C.E., 2005. Remote Sensing of Organic Matter in Coastal Waters, in: Miller, R.L., Del Castillo, C.E., Mckee, B.A. (Eds.), Remote Sensing of Coastal Aquatic Environments: Technologies, Techniques and Applications, Remote Sensing and Digital Image Processing. Springer Netherlands, Dordrecht, pp. 157–180. https://doi.org/10.1007/978-1-4020-3100-7_7

IOCCG (2018). Earth Observations in Support of Global Water Quality Monitoring. Greb, S., Dekker, A. and Binding, C. (eds.), IOCCG Report Series, No. 17, International Ocean Colour Coordinating Group, Dartmouth, Canada.

Pahlevan, N., Mangin, A., Balasubramanian, S.V., Smith, B., Alikas, K., Arai, K., Barbosa, C., Bélanger, S., Binding, C., Bresciani, M., Giardino, C., Gurlin, D., Fan, Y., Harmel, T., Hunter, P., Ishikaza, J., Kratzer, S., Lehmann, M.K., Ligi, M., Ma, R., Martin-Lauzer, F.-R., Olmanson, L., Oppelt, N., Pan, Y., Peters, S., Reynaud, N., Sander de Carvalho, L.A., Simis, S., Spyrakos, E., Steinmetz, F., Stelzer, K., Sterckx, S., Tormos, T., Tyler, A., Vanhellemont, Q., Warren, M., 2021. ACIX-Aqua: A global assessment of atmospheric correction methods for Landsat-8 and Sentinel-2 over lakes, rivers, and coastal waters. Remote Sensing of Environment 258, 112366. https://doi.org/10.1016/j.rse.2021.112366

Valerio, A. de M., Kampel, M., Vantrepotte, V., Ward, N.D., Sawakuchi, H.O., Less, D.F.D.S., Neu, V., Cunha, A., Richey, J., 2018. Using CDOM optical properties for estimating DOC concentrations and pCO 2 in the Lower Amazon River. Optics Express 26, A657. https://doi.org/10/gd8zmb

The use of drones to monitor water quality in inland, coastal and transitional waters is relatively new. The technology can be seen as complementary to satellite and in-situ observations. While cloud cover, low revisit times or insufficient spatial resolution can introduce gaps in satellite-based water monitoring programs, airborne drones can fly under clouds at preferred times capturing data at cm-resolution. In combination with in-situ sampling, drones provide the broader spatial context and can collected information in hard-to-reach areas.

Although drones and lightweight cameras are readily available, deriving water quality parameters is not so straightforward. It requires knowledge of the water optical properties, the atmospheric contribution and special approaches for georeferencing of the drone images. Compared to land applications, the dynamic behavior of water bodies excludes the presence of fixed reference points, useful for stitching and mosaicking and the images are sensitive to sun glint contamination. We present a cloud-based environment, MAPEO-water, to deal with the complexity of water surfaces and retrieve quantitative information on the water turbidity, the chlorophyll content and the presence of marine litter/marine plastics.

MAPEO-water supports already a number of camera types and allows the drone operator to upload the images in the cloud. MAPEO-water also offers a protocol to perform the drone flights and allow efficient processing of the images from raw digital numbers into physically meaningful values. Processing of the drone images includes direct georeferencing, radiometric calibration and removal of the atmospheric contribution. Final water quality parameters can be downloaded through the same cloud platform. Water turbidity and chlorophyll retrieval are based on spectral approaches utilizing information in the visible and Near Infrared wavelength ranges. Drone data are complementary to both satellite and in-situ data. Marine litter detection combines spectral approaches and Artificial Intelligence. Showcases including satellite, drone and in-situ observations will demonstrate the complementary of all three techniques.

WQeMS is a consortium of 11 partners spread all over Europe: Centre for Ecological Research and Forestry Applications (CREAF) (Spain), EOMAP GMBH & CO KG (EOMAP ) (Germany), Cetaqua, Centro Tecnológico del Agua, Fundación Privada ( CETAQUA) (Spain), Autorita' Di Bacino Distrettuale Delle Alpi Orientali ( AAWA ) (Italy), Serco Italia SpA (SERCO) (Italy), Thessaloniki Water Supply and Sewerage Company SA (EYATH SA) (Greece), Engineering - Ingegneria Informatica S.p.A (ENG) (Italy), Finnish Environment Institute (SYKE) (Finland), Phoebe Research and Innovation Ltd (PHOEBE) (Cyprus), Empresa Municipal de Agua y Saneamiento de Murcia, S.A ( EMUASA) (Spain). These organizations cooperate to offer cutting-edge EO technology by creating the ‘Copernicus Assisted Lake Water Quality Emergency Monitoring Service’ (WQeMS) Research and Innovation Action H2020 project. WQeMS aims to provide an open surface Water Quality Emergency Monitoring Service (https://wqems.eu/) to the water utilities’ industry leveraging on the Copernicus products and services. Target is the optimization of the use of resources by gaining access to frequently acquired, wide covering and locally accurate water-status information. Citizens will gain a deeper insight and confidence for selected key quality elements of the ‘water we drink’, while enjoying a friendlier environmental footprint.
There are four services offered by WQeMS. Two services are related to slow developing phenomena, such as geogenic or anthropogenic release of potentially polluting elements through the bedrock or pollutants’ leaching in the underground aquifer through human activities. The other two services are related to fast developing phenomena, such as floods spilling debris and mud or chemicals/ oils spills or algal bloom and potential release of toxins by cyanobacteria at a short time interval bringing sanitation utilities at the edge of their performance capacity. Furthermore, an alerting module is being developed that will deliver alerts about incidents, derived from the WQeMS and Twitter data harvesting module, while a training set of activities will allow users and end-users to gain insight and familiarity with Copernicus Services and the ability to understand and use the functionality of the WQeMS.
The WQeMS system will enable the optimization of the use of resources by gaining access to frequently acquired, wide covering and locally accurate water-status information. WQeMS will generate knowledge that shall support existing decision support systems (DSSs) in a syntactically and semantically interoperable manner. A wide set of parameters will be provided that are useful for the quality assessment of raw drinking water, as captured by existing and emerging requirements of the water utilities industry. It will be based on a modular architecture composed of three main layers: frontend, middleware and backend. Adding to the aforementioned, it will promote further alignment of existing decision support and implementation chains with the updated Drinking and Water Framework Directives.
WqeMS relies on the Copernicus Data and Information Access Services (e.g. DIAS ONDA) for data provision; aiming also at connection with further exploitation platforms. The overall system will be hosted on the ONDA DIAS Cloud infrastructure and will be designed as a microservice, container-based architecture. The cloud nature of the platform, together with the microservice architecture of the solution, provide multiple benefits to the critical objective of the current project, such as (i) Data availability, (ii) Fault Tolerance, (iii) Data Interoperability and (iv) Scalability. The decision to host WQeMS on ONDA will allow having location-based applications taking advantage of proximity of the data.
The objective is to generate an outcome at the end of the project that best suits the interests of the users and citizens, while also enabling compatibility, synergy and complementarity with existing infrastructure and services. Main ambition is to receive approval by the Member States to be embedded in the existing Copernicus Services portfolio. Activities and results are expected to contribute to Europe's endeavors towards GEO and its priorities in the framework of the UN 2030 Agenda for Sustainable Development, the Paris Climate Agreement and the Sendai Framework for Disaster Risk Reduction. WQeMS components, structure and progress are presented and discussed.
This project has received funding from the European Union’s Horizon 2020 Research and Innovation Action program under Grant Agreement No 101004157

Dissolved organic matter (DOM) is important for the function of aquatic ecosystems and can be used as a representation of the lake’s metabolome. DOM can enter an aquatic system via the runoff of the rainfall (or melting tundra) over the ecosystem’s watershed or from in water algal or microbial production. DOM optically detectable fraction – the colored dissolved organic matter (CDOM) – is often used as a proxy for dissolved organic matter. CDOM absorbs radiation in the ultraviolet and visible region of the spectrum and can be identified from satellite imagery. It originates from the degradation of plant materials and other organisms or from terrestrially imported substances. The source of CDOM is important information to understand environmental-driven dynamics in aquatic systems. Fluorescence spectroscopic techniques, such as the excitation–emission matrix (EEM) and the parallel factor analysis (PARAFAC), have been used to distinguish between allochthonous (humic-like) and autochthonous (protein-like) sources. Here, we assess the relationship between these fluorescent components and optical properties such as remote sensing reflectance and inherent optical properties (IOPs) in 19 lakes located within the Mecklenburg–Brandenburg Lake district in the North German Lowland. These lakes differ in size, shape, depth, trophic state and biogeochemical characteristics. Most lakes are connected in-series by rivers and natural or manmade channels. Water samples from these lakes were analyzed for absorbance and fluorescence measurements by using spectrophotometers. These samples were also used for the computation of the absorption coefficients of phytoplankton, CDOM, and non-algal particles, which were used as the IOP dataset. Remote sensing reflectance was calculated from radiometric measurement at the water surface using two handheld spectroradiometers (ASD, JETI). We started calculating 2 PARAFAC components yielding in a high correlation between both (Spearman’s rs of 0.88) indicating that it is difficult to differentiate these two components. Calculating 4 PARAFAC components, we observed a high correlation between component 1 and 2 (Spearman’s rs of 0.98) and between component 1 and 3 (Spearman’s rs of 0.87). Component 4 was the least correlated with Spearman’s rs of 0.23, 0.25 and 0.09 with components 1, 2 and 3 respectively. This indicates that it would be possible to differentiate components 1, 2 and 3 from component 4. The 2-dimensional correlation plot with the remote sensing reflectance and each component showed that for components 1, 2 and 3 the reflectance ratio at wavelengths 620 nm/590 nm was the most appropriate, while for component 4 the reflectance ratio was most appropriate at 825 nm/665 nm. In relation to the IOPs, the correspondence analysis showed that components 1, 2 and 3 are related to the absorption coefficient of CDOM and the absorption coefficient of non-algal particles while component 4 was related to the absorption coefficient of CDOM and the absorption coefficient of phytoplankton. These results indicate that components 1, 2 and 3 are related to the allochthonous CDOM while component 4 seems to be related to the autochthonous CDOM. Additionally, it also shows the potential of remote sensing for the identification CDOM sources which can help to understand aquatic ecosystem dynamics under environmental change.

MAPAQUALI - Customizable modular platform for continuous remote sensing monitoring of aquatic systems

The water resource is vital, not only for maintaining life on Earth, but also for supporting economic development and social well-being as the sustainable growth of all the nations depends upon water availability. Approximately 12% of the planet's surface fresh water available for use circulates through the Brazilian territory. Due to this water availability, Brazil has an extensive number of large artificial and natural aquatic ecosystems. This water availability places Brazil in a privileged position, but it also poses a great challenge for sustainable use and monitoring of these natural resources. For instance, the nutrient inflows to lakes and hydroelectric reservoirs from irrigated agriculture and sewage from nearby cities significantly contribute to the eutrophication process and the systematic occurrence of cyanobacterial blooms. These blooms can be harmful and produce toxins that lead to a series of public health problems. Even when not harmful, they impair fisheries and the recreational use of those water bodies. These environmental impacts on aquatic ecosystems need to be determined and monitored, mainly in reservoirs, as energy sources, besides being renewable, must be clean. This study summarizes the integrated effort of specialists in hydrological optics, aquatic remote sensing, and computer science to build a customizable modular platform, named MAPAQUALI. The platform allows a continuous monitoring of aquatic ecosystems based on satellite remote sensing, and the integration of bio-optical models derived from in-situ measurements. The platform will generate and make available, for aquatic ecosystems for which it is customized, a spatiotemporal information about water quality parameters: Chlorophyll-a, Cyanobacteria, Total Suspended Solids, Secchi disk depth, diffuse attenuation coefficient (Kd), and bloom events alerts (especially cyanobacteria).

The MAPAQUALI platform comprises the following modules: Data Pre-processing; Bio-optical Algorithms; Query and View WEB.

The Data Pre-processing Module (DPM) generates and catalogs Analysis Ready Data (ARD) [10.1109/IGARSS.2019.8899846] collections, which are input data for the Bio-optical Algorithms Module (BAM) for water quality products generation. The DPM has data acquisition, processing and cataloging functionalities. The DPM structure is flexible for adding new processing tasks or even new functionalities. The following processing tasks are available in the current implementation of MAPAQUALI: query and image acquisition from data providers (Google Cloud Platform or Brazil Data Cube Platform [10.3390/rs12244033]); atmospheric correction procedure with 6SV model; water bodies’ identification and extraction; cloud and shadow masking; sunglint and adjacency corrections. The BAM comprises parameterized/calibrated/validated algorithms using Brazilian inland water in situ bio-optical datasets (LabISA – INPE), and OLI, MSI, and OLCI simulated spectral bands. Algorithms were parameterized for OLCI sensor only for aquatic systems having the suitable size, such as large lakes in the Amazon floodplain. In addition, to ensure the best possible accuracy, we developed a semi-analytical [10.1016/j.isprsjprs.2020.10.009; 10.3390/rs12172828], hybrid [10.3390/rs12010040], machine learning [10.1016/j.isprsjprs.2021.10.009], and empirical [10.3390/rs13152874] algorithms, using in situ data representative of the full range variability of the apparent and inherent optical proprieties. These algorithms are achieving accurate results. For example, the hybrid algorithms for Chl-a have an error of 20% (MAPE = 20%). Machine learning algorithms, for estimating water transparency, presented errors of approximately 25%. Moreover, Kd algorithm for oligotrophic reservoir resulted in errors of 20%. The Query and View WEB is a web portal providing resources for searching for the aquatic systems integrated into the platform and returns the products available for each of them.

Tools for viewing and analyzing time series are available in this module. Registered users to the platform can freely download products and images from the ARD archive. Additionally, users can consume any available data through our web geoservices enabled database, such as Web Map Service (WMS) or Web Feature Service (WFS).

For the integrated application of DPM and BAM processing tasks, we are using the process orchestration infrastructure available on the Brazil Data Cube Platform [10.3390/rs12244033]. In this way, MAPAQUALI platform can perform all operations periodically, which allows continuous monitoring of the aquatic systems under consideration. At the end of each execution, the newly generated data products are cataloged, and made available for consulting in monitoring activities.

In our ongoing efforts, we are customizing water quality bio-optical algorithms to four aquatic ecosystems: two multi-user reservoirs, a set of lower amazon floodplain lakes, and one nearshore coastal water. As modular and customizable, others aquatic ecosystems can be easily inserted into the platform.

Validating water quality model applications is challenging due to data gaps in in-situ observations, especially in developing regions. To such a challenge, remote sensing (RS) has provided an alternative to monitor the water quality of inland waters due to its low cost, spatial continuity and temporal consistency. However, limited studies have exploited the option of validating water quality model outputs with RS water quality data. With sediment loadings regarded as a threat to the turbidity and trophic status of Lake Tana in Ethiopia, this study aims at using existing RS lake turbidity data to validate the seasonal and long-term trends of sediment loadings in and out of Lake Tana. A hydrologically calibrated SWAT+ model is used to simulate river discharge and sediment loadings flowing in and out of Lake Tana basin. Together, with a remote sensing dataset of lake turbidity from Copernicus Global Land Service (CGLS), seasonal and long term correlations between lake turbidity and sediment loadings at the river mouths of Lake Tana are estimated.
Results indicate a strong positive correlation between sediment load from inflow and out flow rivers with RS lake turbidity (r2 > 0.7). Other strong positive relations were observed between the stream flow from inflow rivers and the lake turbidity (r2 > 0.5). These indicate that river streamflow accounted for significant responses in river sediment loads and lake turbidity which likely occurred from a combination of overland transport of sediment into streams due to erosion of the landscape, scouring of streambanks, and resuspension of sediment from channel beds. We conclude that RS water quality products can potentially be used for validating seasonal and long term trends in simulated SWAT+ water quality outputs, especially in data scarce regions.

Satellite retrieval and validation of bio-optical water quality products in Ramganga river, India
Veloisa Mascarenhas1*, Peter Hunter1, Matthew Blake1, Dipro Sarkar2, Rajiv Sinha2, Claire Miller3, Marion Scott3, Craig Wilkie3, Surajit Ray3, Andrew Tyler1
* veloisa.mascarenhas@stir.ac.uk

1University of Stirling, UK
2Indian Institute of Technology Kanpur, India
3University of Glasgow, UK

In addition to water resources, inland waters provide diverse habitats and ecosystem services. They are threatened however, by unregulated anthropogenic activities and so effective management and monitoring of these vital systems has gained increasing attention over the recent years. Being optically complex waters, inland water remote sensing continues to face challenges underpinning the retrieval of physical and biogeochemical properties. We present here the retrieval and assessment of satellite derived L2 bio-optical water quality products from Sentinel2 and Planet satellites for a highly turbid river system. Bio-optical water quality products including remote sensing reflectance, total suspended matter and chlorophyll-a (Chl-a) concentrations are validated using in situ observations along the river Ramganga, in India. The Ramganga has a large (22,644 km2) diverse catchment, with intensive agriculture, extensive industrial development and a rapidly growing population. The over-abstraction of both surface and groundwater, and pollution due to industrial and domestic waste, mean the Ramganga presents an ideal case study to demonstrate the value of satellite data for monitoring water quality in a highly impacted river system. For the case study, five different atmospheric correction methods are tested in processing the Level 1 Sentinel 2 imagery and a set of biogeochemical algorithms to estimate bio-optical products. Additional bio-optical products such as turbidity are estimated from satellite derived remote sensing reflectance to be matched with in situ turbidity observations. The Sentinel dataset is supplemented using high resolution (3-5 m) imagery from commercial satellite, Planet, processed using ACOLITE atmospheric correction method. The river transect is characterised by high variability in optically active constituents and remote sensing reflectance. Around the Moradabad area, in situ measured turbidity values peak during the month of July while Chl-a concentrations are observed to be highest in early May.

Quantum yield of fluorescence (ϕ_F) represents the small fraction of absorbed photons in phytoplankton that is converted to sun-induced fluorescence (SIF). This fraction is typically up to 2% in optically complex waters. All other absorbed photons are either used for photochemistry in the reactions centers or dissipated as heat. When fluorescence is reduced from a maximum level due to an increase in open reaction centers, Photochemical Quenching (PQ) occurs. Other forms of fluorescence reduction lead to increased thermal dissipation and are referred to as Non-photochemical Quenching (NPQ). In cases where NPQ is minimal, ϕ_F and SIF increase with higher irradiance. However, when NPQ is present due to photo-inhibition or protective measures employed by the phytoplankton, SIF may still increase with irradiance while ϕ_F decreases. Consequently, NPQ conditions also lead to lower quantum yield of photosynthesis.

Knowing the ϕ_F is key to understanding SIF emission in phytoplankton as it enables us to interpret the dynamics of SIF in relation to PQ or NPQ. Disentangling PQ from NPQ allows us to use SIF estimates in various applications in aquatic optics and remote sensing such as accurate estimation of chlorophyll-a concentration (chl a) or modelling of primary productivity. These are essential to assess the water quality status of surface waters and to understand the dynamics of aquatic ecosystems. Retrieving and interpreting SIF becomes more plausible at the present time and in the near future with the increasing availability of in-situ, airborne and spaceborne hyperspectral sensors. However, obtaining ϕ_F is challenging due to prior data necessary for the calculations especially in inland waters.

Using the autonomous Thetis profiler from the LéXPLORE platform in Lake Geneva, we demonstrate a novel way of estimating ϕ_F based on an ensemble of in-situ profiles of Inherent Optical Properties (IOPs) and Apparent Optical Properties (AOPs) taken between October 2018 and August 2021. In particular, we exploited the profiler’s hyperspectral radiometers to obtain upwelling radiances and downwelling irradiances in the top 50 m of the water column. These AOPs were the main basis of our SIF retrieval, representing natural variations in fluorescence emission under different bio-geophysical conditions. We further used hyperspectral absorption and attenuation, and backscattering measurements at discrete wavelengths to obtain the water’s IOPs. These IOPs were used in radiative transfer model simulations assuming ϕ_F=0 to obtain a second set of AOPs without fluorescence contributions. The measured and simulated reflectances obtained outside the fluorescence emission region which satisfy the optical closure analysis were kept in the succeeding steps. By associating the difference between these measured and simulated AOPs, known chlorophyll-a concentrations and IOPs, we obtained estimates of ϕ_F.

We analysed obtained ϕ_F values to determine the conditions at which NPQ occurs. Consequently, we evaluated the vertical and temporal changes in ϕ_F. We observed diurnal changes in NPQ occurrence, particularly during clear sky conditions where downwelling irradiance changes significantly throughout the day. For instance, we observed that ϕ_F can be up to 65% lower when NPQ is activated compared to PQ stimulated conditions. While downwelling irradiance is a significant contributor to changes in ϕ_F, its role can be sometimes not easily interpreted because the threshold of radiant flux at which NPQ is activated in inland waters is not consistent. Other factors such as phytoplankton photo-adaptation and the composition of different phytoplankton communities also play significant roles in understanding phytoplankton response to incident light and therefore, quenching mechanisms. Our results contribute insight on the nature of SIF and can facilitate activities to assimilating SIF and ϕ_F estimates in remote sensing algorithms, which would aid us in monitoring not only phytoplankton biomass but also the eco-physiological state of phytoplankton cells.

Algal blooming is one of the factors with the greatest impact on the quality, functioning, and ecosystem services of waterbodies, and can frequently occur in the coastal regions (O'neil et al., 2012). The observed increase in cyanobacterial blooms in European seas is attributed to severe eutrophication and a subsequent change in nutrient balance caused by anthropogenic nutrient enrichment, in particular from urban areas, agriculture and industry (Kahru et al., 2007, Vigouroux et al., 2021). The EU Marine Strategy Framework Directive (MSFD) is the main initiative to protect the seas of Europe that requires to minimize human-induced eutrophication (MSFD, 2008). The majority of indicators developed under MSFD Descriptor 5 Eutrophication are based on in situ monitoring data, and only recently, the Earth-Observation (EO) data has started to be proposed as a valuable source of information for monitoring, ecological status assessment and indicator development (Tyler et al., 2016). Recently, HELCOM proposed a pre-core indicator Cyanobacteria Bloom Index (CyaBI) that evaluates cyanobacterial surface accumulations and cyanobacteria biomass, describes the symptoms of eutrophication caused by nutrient enrichment, and exclusively is based on EO satellite data (Antilla et al., 2018). The indicator was developed using the Baltic Sea as a testing site, and is focused on the open sea areas (HELCOM, 2018). However, anthropogenic pressures, unbalanced and intensive land use, and climate change increasingly affect coastal and transitional waters representing water continuum from inland waters towards sea. These regions are more exposed to ongoing eutrophication and the severe cyanobacteria blooms are evident (Vigouroux et al., 2021). Therefore, the aim of this study is to test the applicability of pre-core indicator CyaBI for the coastal and transitional waters of the two enclosed seas located at different latitudes: the Baltic and the Black Sea. We also hypothesize that the intensive cyanobacteria blooms significantly alter the short-term environmental conditions of the Seas in terms of the Sea Surface Temperature (SST) changes.
The Baltic and the Black Sea are the world’s largest brackish water ecosystems, which exhibit many striking similarities as geologically young post-glacial water bodies, semi-isolated from the ocean by physical barriers. Both Seas are exposed to similar anthropogenic pressures, such as increasing urbanization, water pollution by heavy industries, intense agriculture, overexploitation of fish stocks, abundant sea traffic and port activities, oil spills, etc. In both seas, increasing attention is being paid to the search for scientifically based solutions to improve the state of the marine environment.
In our study, we have used time series of the Medium Resolution Imaging Spectrometer (MERIS) on-board Envisat at 300 m, and the Ocean and Land Colour Instrument (OLCI) on-board Sentinel-3 at 300 m spatial resolution for the estimation of chlorophyll-a (Chl-a) concentration. Chl-a concentration was retrieved after the application of the FUB processor, which was developed by the German Institute for Coastal Research (GKSS), Brockmann Consult, and Freie Universität Berlin, and is designed for European coastal waters. In case of MERIS images, the FUB processor uses Level 1b top-of-atmosphere radiances to retrieve the concentrations of the optical water constituents. A good agreement (R2=0.69, RMSE=14.44, N=56) was found between Chl-a derived from MERIS images after application of FUB processor and in situ measured Chl-a concentration during the validation in the coastal waters of the Lithuanian Baltic Sea (more details in Vaičiūtė et al., 2012). Although the FUB processor is originally designed for MERIS images, we have tested its performance in case of OLCI images. Chl-a concentration derived from OLCI data after FUB processor application and measured in situ were in agreement with R2=0.72, RMSE=4.2, N=31. CyaBI index was calculated following the methodology described in Antilla et al. (2018). In this study, we used Terra/Aqua MODIS standard Level 2 SST products with a spatial resolution of around 1 km—obtained from the NASA OceanColor website—to analyse the spatial patterns and changes in SST at the presence of cyanobacteria surface accumulations.
In this presentation, we will demonstrate the first results of ecological status assessment using pre-core CyaBI indicator in the Lithuanian Baltic and Ukrainian Black Seas. We will discuss the potential of using CyaBI for the ecological status assessment in the coastal and transitional waters, and for the seas located at different latitudes. We also will provide significant insights about the integration of SST data for the ecological status assessment considering the Descriptor 5 of MSFD and the Water Framework Directive.
The research was funded by the Lithuanian-Ukrainian bilateral cooperation in the field of science and technology under project "Measuring the marine ecosystem health: concepts, indicators, assessments – MARSTAT (contract no. S-LU-20-1)".

Monitoring is an integral precondition to determine lakes' ecological status and develop solutions to restore lakes that have deteriorated from reference conditions. Spatial and temporal limitations of conventional in situ monitoring impede adequate evaluation of lakes' ecological status, especially when dealing with large-scale measurements. Sentinel-2 (S2) - a constellation of the two twin satellites, S2-A and S2-B with the MultiSpectral Instrument (MSI) on board can be a complement to in situ data. S2 MSI imagery makes possible investigation of even small water bodies due to its high spatial resolution of 10, 20 and 60 meters depending on the spectral band. Besides, S2 spectral resolution allows estimation of a wide range of water quality parameters such as chlorophyll-a (chl-a), water color, colored dissolved organic matter (CDOM), etc. However, implementing the remote sensing data for water quality assessment over small inland waters might be obstructed by the adjacency effect (AE). AE is especially strong in small, narrow, or complex-shape water bodies surrounded by dense vegetation and decreases further offshore. Therefore, the largest possible homogeneous water area surrounding the sampling point would increase the possibility to obtain an accurate signal from the water’s surface called water-leaving reflectance ρω(λ). Moreover, a combination of chl-a, CDOM and TSM concentrations also affect the probability and accuracy of ρω(λ) and must be considered.
Test sites of this study are optically complex lakes of Northern Europe with a high and varying amount of optically active substances. In this study, the dataset of 476 in situ measurements of water properties from 44 lakes were used. Measured concentrations of chl-a are ranged between 2 mg/m3-100 mg/m3, total suspended matter (TSM) 0.6 mg/m3 - 48 mg/m3 and aCDOM (442) 0.5 – 48 m-1. Water-leaving reflectance ρω(λ) was measured deploying above-water RAMSES TriOS radiometers.
The aim of this study was to evaluate the capabilities and limitations of the S2 MSI data after atmospheric correction by POLYMER 4.12 and C2RCC v1.5 processors. The results were analysed together with lakes’ area and shape complexity (shape index, SI) and the signal strength as determined by the concentration of chl-a, TSM and aCDOM.
The objectives of the study were:
1. Validate and analyze POLYMER and C2RCC-derived ρω(λ) against in situ measurements using match-up analysis for exact location (1 x 1), 3 x 3 and 5 x 5-pixel size region of interest (ROI).
2. Evaluate spatial distribution and homogeneity of POLYMER and C2RCC quality flags and water quality products. Based on that derive area and SI thresholds of the lakes that can be monitored with S2 (20 m spatial resolution).
3. Evaluate the spatial and temporal distribution of the failures in POLYMER and C2RCC atmospheric correction and in the resulting water quality maps. Analyze its impact on the derived ecological status class in optically different lakes.
The validation of POLYMER ρω(λ) product against in situ measurements resulted in slightly better accuracy than the C2RCC product. For the bands at 560 nm, 665 nm and 705 nm, crucial to derive chl-a over optically complex waters, POLYMER showed a weak correlation (R2 = 0.41, 0.12, 0.36) for 1 x 1 area, however, R2 for the 3 x 3 region was higher and equaled 0.63, 0.48 and 0.58, respectively. Noticeably, with enlarging the ROI up to 5 x 5 pixels grid, R2 decreased and equaled 0.36, 0.33 and 0.31 for 560 nm, 665 nm and 705 nm, respectively, which indicates nonhomogeneity in pixels distribution. Moreover, 5 x 5 ROI is 10000 m2 area, which might be too large to compare with the field measurement from only one point. The coefficient of determination for C2RCC data increased with the enlargement of the ROI to a 3 x 3 and 5 x 5-pixel area similar to POLYMER, however, not so noticeably. Specifically, for exact location (1 x 1 ROI), R2 equaled 0.45, 0.35, 0.40 at 560 nm, 665 nm and 705 nm wavebands, whereas for 3 x 3 and 5 x 5 pixel area it equaled 0.48, 0.41, 0.40 and 0.50, 0.38, 0.40, respectively.
POLYMER quality flags of the S2 imageries sensed in spring, summer and autumn over a group of 1727 lakes predominantly located in Southern Estonia were analyzed. In spring, most of the water bodies under 1 ha did not have valid quality flags. Besides, more complex-shape water bodies (SI > 2) with no valid quality flags were even larger (up to 6 ha). It was shown that the amount of quality flags, usable to produce water quality maps, decreases towards autumn.
Spatial and seasonal evaluation of the chl-a was conducted in the optically and geometrically different lakes. Failures in POLYMER atmospheric correction resulting in abnormal chl-a values were mostly due to the combined effect of optical properties of the water bodies and adjacency effect, strongest over clear waters surrounded by forest. This resulted in very few pixels and also with high spatial heterogeneity over clear water lakes. Whereas over eutrophic waters, there were more quality controlled satellite retrievals with improved spatial patterns of chl-a. It was shown that S2 MSI is a promising method for studying water bodies but the adjacency to the shore and the level of optically active substances must be considered.

Remote sensing-based products are widely used for scientific research and synoptical monitoring of water resources. The use of satellite-based products provides a less costly and time-consuming alternative to traditional in-situ measurements. The conservation of water resources poses a challenge on multiple levels, including local institutions, authorities, and communities. Therefore, the monitoring of water resources, in addition to provide a scientific output, should devote its efforts also to the publication and sharing of the results. Therefore, communication, coordination, and publishing of data are essential for preserving the water ecosystems.
This work presents the design and implementation of two components of the IT infrastructure for supporting the monitoring of lake water resources in the Insubric area for SIMILE ("Integrated monitoring system for knowledge, protection and valorisation of the subalpine lakes and their ecosystems"; Brovelli et al., 2019) Italy-Switzerland Interreg project. SIMILE monitoring system benefits from various geospatial data sources such as remote sensing, in-situ high-frequency sensors, and citizen science. The infrastructure uses and benefits from Free and Open-Source Software (FOSS), open data and open standards, facilitating the possibility of reuse for other applications.
The designed applications aim at enhancing the decision-making process by providing access to remote-sensing based lake water quality parameters maps produced under the project for Lakes Maggiore, Como and Lugano. The satellite monitoring system for SIMILE considers estimating different water quality parameters (WQP) using optical sensors. The analysed water quality parameters maps include the concentration of Chlorophyll-a (CHL-a), Total Suspended Matter and Lake Surface Water Temperature (LSWT). Each product is delivered with a specific spatial and temporal resolution depending on the sensor used for the monitored parameter. The WQPs maps production frequency is affected by factors such as the revisit time of the sensor over the study area and the cloud coverage. CHL and TSM are monitored with the ESA Sentinel-3A/B OLCI (Ocean and Land Colour Instrument) whose spectral bands include the visible and infrared portions of the spectrum. LSWT is monitored using the NASA Landsat 8 TIRS (Thermal Infrared Sensor). Sentinel-3 A/B offers a daily revisit time over the study area with a resolution of 300m, which, on average, allows for the production of CHL-a and TSM maps weekly. Landsat-8 satellite provides a higher spatial resolution of 30m, but with a revisit time of 16 days, then, on average, allows for the production of LSWT maps monthly.
The archiving and sharing of the WQPs maps are of interest to the SIMILE project. In particular, the project promotes the publication of the data as time series to monitor the evolution of the different WQP maps. WQP maps can support the assessment of various processes taking place inside the aquatic ecosystems, for example, the eutrophication level in a water body from CHL-a. Sediment concentration, which can be deduced from TSM maps can influence the penetration of light, ecological productivity, and habitat quality, and can harm aquatic life. LSWT maps allow exploring lake dynamics processes such as sedimentation, concentration of nutrients and the presence of aquatic life, but also the temporal variability of temperature due to climate change (Lieberherr et al, 2018).
Two web applications have been designed aiming at simplifying the data-sharing process and allowing for the interactive visualization of the WQPs maps. The first one is built on GeoNode, to upload, edit, manage and publish the WQP maps. GeoNode is an open-source Geospatial Content Management System that eases the data-sharing procedures. The second one, is a WebGIS application that aims at providing a user-friendly environment to explore the different WQPs maps. The WebGIS benefits from OGC standards, such as the Web Mapping Service (WMS), to retrieve and display the maps published on the GeoNode application. The publication of the datasets through OGC standards is possible thanks to the GeoServer instance working on the back-end of the GeoNode project.
SIMILE WebGIS goal is favouring the visualization and query of lakes WQP as time series. For this reason, it was possible to exploit the raster data format support available into the data-sharing platform. Indeed, GeoNode permits the upload of raster data in GeoTIFF format, taking advantage of the data storage system implemented by GeoServer. Note that GeoServer provides additional multidimensional raster data support (such as image mosaics and NetCDF), which enables the storage of the collection of datasets with a time attribute. Nonetheless, GeoNode does not support the multidimensional raster data formats, and using them would imply the need of direct interaction with the remote server hosting GeoServer. The interaction with the remote server represents a barrier to the data sharing workflow (due to additional File Transfer Protocols to send the data to the server). The GeoTIFF format does not provide a time attribute. In order to overcome this limitation and allow the management of time series, a naming convention has been introduced and the timestamp is provided in the layer name. Next, for matching layer typologies, it was possible to build groups of layers by extracting unique dates values. The constructed layers groups used the collection event of "LayerGroups" for the "Layer" object in OpenLayers Library. Thus, the time series visualization for WQP in the WebGIS was possible while maintaining GeoNode as a suitable tool for the publication of raster data.
Therefore, the WQP maps are provided with a naming convention which describe the sensor used for the acquisition, the product typology, the coordinate reference system of the map, and the timestamp of the image acquisition, in order to facilitate the integration of maps in the database and the metadata compilation. An example of the naming convention is “S3A_CHL_IT_20190415T093540”. Here, the file name contains information corresponding to the coordinate reference system (“IT”, WGS84 – UTM32N), the sensor involved in the acquisition of the imagery (“S3A”, ESA Sentinel3A-OLCI), the product’s typology (“CHL”, Chlorophyll-a), and the timestamp of the retrieval of the imagery (“20190415T093540”, April 15, 2019, at 09:35:40), all separated by an underscore. The application has been designed to let web client users display the layers in time, taking advantage of the map timestamp. Moreover, the naming convention supported the styling of the layers and the metadata preparation and display.
The WebGIS builds upon a node.js runtime environment that allows creating server-side applications using JavaScript. The WebGIS design benefits from the OpenLayers and JQuery JavaScript libraries and the VueJS framework. Accordingly, the web application integrates capabilities and tools which are built using components that can be attached/detached from the application if needed. The WebGIS components, hereafter panels, include a Layer Panel, a Metadata Panel, a Time Manager Panel and a BaseMap Panel. The different Panels will be populated by parsing the information obtained from the WMS getCapabilities operation from GeoServer. The Layer Panel integrates the list of layers available in GeoNode. Each item in the list of layers allows users to control the visibility of the layers (i.e., display and opacity), download the datasets and explore the metadata (for a selected layer). The Metadata Panel includes an abstract according to the layer typology, the start/end dates for the first/last map, and the symbology to describe the corresponding layer. In addition, the Metadata Panel makes use of the getLegendGraphic operating to retrieve the layer legend. The Time Manager Panel contains controllers that enable the querying and visualization of raster time series. At last, the BaseMap Panel provides various options for changing the base map of the WebGIS.
The web-based application implemented in this work provides a mechanism for sharing and monitoring water quality parameters maps. The infrastructure implements two different applications focusing on two different audiences. First, the collaborative data-sharing platform (GeoNode) that targets the map producers allowed to upload and manage the lake water quality maps (following the naming convention for the products). Second, the WebGIS aims at becoming an open application for the exploration of the products uploaded into the GeoNode platform. The WebGIS provides an interactive application to display the lake water quality products as time series in a user-friendly environment. The components inside the WebGIS provide users to control the visibility of the layers, query maps in time, explore the layers metadata and customize the base map background. Data accessibility for water quality parameters enables the monitoring and assessment of the water bodies health. Moreover, the monitoring of the water resources is mandatory for guaranteeing the livelihood of the nearby communities depending on its consumption and quality.

Brovelli, M. A., Cannata, M., & Rogora, M. (2019). SIMILE, A GEOSPATIAL ENABLER OF THE MONITORING OF SUSTAINABLE DEVELOPMENT GOAL 6 (ENSURE AVAILABILITY AND SUSTAINABILITY OF WATER FOR ALL). ISPRS - International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLII-4/W20, 3–10. https://doi.org/10.5194/isprs-archives-XLII-4-W20-3-2019

Lieberherr, G.; Wunderle, S. Lake Surface Water Temperature Derived from 35 Years of AVHRR Sensor Data for European Lakes. Remote Sens. 2018, 10, 990. https://doi.org/10.3390/rs10070990

The thaw lakes and drained thaw lake basins are a prominent feature in the Arctic and cover large areas of the landscapes in the high latitudes. Thaw lakes as well as drained thaw lake basins have major impacts on a region’s hydrology, landscape morphology and flora and fauna. Drained lake basins have been studied across regions in the Arctic and differences in abundance and distribution exist between regions in the circumpolar Arctic. Thawing permafrost and drainage lakes can also affect human activities. Our research area is the Yamal peninsula in the Western Siberia, Russia. Yamal Peninsula is about 700 km long and about 150 km wide, extends from 66° to 72° degrees North. In Yamal petroleum industry with related infrastructure networks can be affected by changes in lake and stream hydrology. Nenets reindeer herding is the traditional land use form in the Yamal. Reindeer herding is based on natural pastures and resources, and lakes and streams serve as important fishing resource for own use and sale. Thawing and drained lakes are part of climate change driven landscape changes in the area.
Landsat has been used in multiple studies for the analysis of lake area, extent and drainage or shrinkage events in the circumpolar Arctic. To analyze lake drainage, lake shrinkage and changes in lake extent consistent satellite data with adequate temporal and spatial resolution is needed. Frequent cloud cover in arctic regions during the summer months limits the number of suitable acquisitions of multispectral sensors and hinders the implementation of large-scale time series analysis efforts. Landsat data enables time span from 1972, although Landsat MSS images were rather coarse and good quality images are sparse. Good quality data have only been available since the mid-1980s when Thematic Mapper was launched. Old archival aerial photographs allows looking further back in time, in some cases even to 1940’s, but their limited spatial coverage and availability does enable large scale investigations. Cold war era spy satellite missions like Corona and KH are the only options to expand time span to late 1950’s and early 1960’s.
Our remote sensing datasets cover a period from 1961-2019. Corona data represents the oldest data source and mosaic was compiled from 38 original Corona images. Corona mosaics resolution is about 7 meters. Landsat mosaics are derived from 1980’s and 2010’s data. In addition we use several very high resolution satellite data (Quickbird-2, Worldview-2/3) and drone data to demonstrate lake changes in detail. Field data for verification of drained lakes have been collected from several parts of Yamal. Field data includes observations of changes and vegetation sampling. We have also interviewed several reindeer herders to understand the implications of lake changes for reindeer husbandry.
The changes were observed in the following periods 1961-1988 and 1988-2018. The results show that the disappearance of the lakes occurs throughout the period, but in the latter period the process accelerated. In terms of reindeer husbandry, the issue is multidimensional, as lakes that are quite important for fishing have disappeared in some places. Drained Lake, on the other hand, will soon turn into a good quality pasture land where nutritious grasses and forbs grow, but if drained lake is located in a winter grazing area, it is only a lost fish resource.

Figure: On the left, a partially drained lake, about 10 years ago. Old lake bottom grows with dense grass, sedge and forb carpet. On the right about 2-3 years ago partially drained lake, the revegetation is much slower which is also due to the sandy soil.

Retrieving sea-surface salinity near the sea-ice edge using spaceborne L-band radiometers (SMOS, Aquarius, SMAP) is a challenging task. There are several reasons for this. First, in cold water, the sensitivity of the L-band emitted surface brightness temperature to salinity is small, which results in large retrieval errors. Additionally, it is diffi-cult to both detect the sea-ice edge and accurately measure small sea-ice concentrations near the sea-ice edge. We have evaluated several publicly available sea-ice con-centration products (OSI-SAF, NSIDC CDC, NCEP) and found that none of them meet the accuracy that is required to use them as ancillary input for satellite salinity retriev-als. This constitutes a major obstacle in satellite salinity measurements near the sea-ice edge. As a consequence, in the cur-rent NASA/RSS V4.0 SMAP salinity release, salinity cannot be retrieved over large areas of the polar oceans.
We have developed a mitigation strategy that directly uses AMSR2 TB measurements of the 6 – 36 GHz channels to assess the sea-ice contamination within the SMAP antenna field of view instead of external ancillary sea-ice concentration products. The 6 and 10 GHz AMSR2 TB show very good correlation with the SMAP TB when averaged over several days. Moreover, the spatial resolutions of AMSR2 6 GHz and SMAP TB measurements are very comparable.
Based on this, we have developed a machine-learning algorithm that uses the AMSR2 and SMAP TB as input (1) to detect low sea-ice concentrations within the SMAP foot-print and (2) that allows to remove the sea-ice fraction from the SMAP measurements.
This new algorithm allows for more accurate sea-ice detection and mitigation in the SMAP salinity retrievals than the ancillary ice products do. In particular, the detection of icebergs in the polar ocean and salinity retrievals in the vicinity of sea-ice can be significantly improved. We plan to apply this new method in the upcoming NASA/RSS V5.0 SMAP salinity release.

Understanding surface processes during sea ice melt season in the Arctic Ocean is crucial in the context of ongoing Arctic change. The Chukchi and the Beaufort Seas are the Arctic regions where salty and warm Pacific Water (PW) flows in from the Bering Strait and interacts with sea ice, contributing to its melt during summer. For the first time, thanks to in-situ measurements from two saildrones deployed in summer 2019, and SMOS (Soil Moisture and Ocean Salinity) and SMAP (Soil Moisture Active Passive) satellite Sea Surface Salinity (SSS), we observe large low SSS anomalies induced by sea ice melt, referred as meltwater lenses (MWL).

The largest MWL observed by the saildrones during this period covers a large part of the Chukchi shelf. It is associated with a SSS anomaly reaching 5pss, and persists for a long time (up to one month). In this MWL, the low SSS pattern influences the air-sea momentum transfer in the upper ocean, resulting in a reduced shear of currents between 10 and 20 meters depth.

L-Band radiometric SSS allows an identification of the different water masses found in the region during summer 2019 and their evolution as the sea ice edge retreats over the Chukchi and the Beaufort Seas. Two MWL detected in these two regions exhibit different mechanism of formation: in the Beaufort Sea, the MWL tends to follow the sea ice edge as it retreats meridionally whilst in the Chukchi Sea, a large persisting MWL generated by the advection of a thin sea ice filament is observed.

Taking advantage of the demonstrated ability of SSS satellite observations to monitor MWL and the 12 years-long SMOS time series, we further examine the interannual variability of SSS during sea ice retreat over the Chukchi and Beaufort Seas for the last 12 years.

The aim of work is evaluation phytoplankton’s influence on the carbon cycle, oxygen concentration, and ocean dwellers food web in the Greenland Sea.
There are several tasks to achieve our goal: 1) to study the interaction between chlorophyll and physical properties of the sea water; 2) to determine seasonal cycle of spatial pattern and vertical profile of phytoplankton; 3) to estimate primary production.
To begin with, the Arctic waters are quite unstable in terms of sea ice thickness, open water area, research accessibility, and, moreover, are likely to face ice-free summers in the near future. As a result, these changes provide changes of the light absorbance, nutrient distribution and phytoplankton seasonality. And yet, it is still unknown if it decreases or increases phytoplankton primary production in the Greenland Sea. There is a lack of field data, hence, satellite data provide an alternative.
Phytoplankton are responsible for releasing half of the World’s oxygen and over 90% of marine primary production. Our work combines satellite and field data to investigate seasonal cycle, variability and productivity of phytoplankton in the Greenland Sea (Fram Strait), and apply modelling techniques to estimate primary production of the area.
Satellite HERMES GlobColor data were processed by MatLab/Python. Field data was used to recover Gaussian coefficients in order to apply them to the satellite data, what gives us possibility to implement depth profiles and establish Euphotic depth for every data cell.
From the field data from the Fram Strait’s 2021 RV Kronprinz Haakon summer cruise chlorophyll-a, light absorbance by particulates and primary production were measured. Using remote sensing data of chlorophyll concentration, sea temperature and photosynthetically available radiation, we were able to obtain modelled estimations of primary production. Further, field and satellite data primary production estimations were compared.
We plan that our primary production estimates will be used for validation of other biogeochemical models.

The inter-annual changes of the Arctic Ocean (e.g. dense water formation, meridional heat redistribution) are well-known proxies of global climate change. The ocean circulation in the high latitude seas and Arctic Ocean has significantly changed during recent decades, with significant impact on the socio-economic activities for the locals. Monitoring the Arctic environment is however non-trivial: the Arctic observing network is notably lacking the capability to provide a full picture of the ocean variability due to technological and economical limitations to sample the seawater beneath the sea ice or in the marginal ice zones. This leads to the obvious need of optimizing the exploitation of data from space-borne sensors. For more than two decades, altimetric radars measuring the sea level at millimetric precision have revolutionized our knowledge of the global mean sea level rise and oceanic circulation. Technological solutions are continuously needed and pursued to enhance the spatial resolution of the altimetric signal and enable the solution of the mesoscale dynamics, either in the design of the altimeter itself (e.g. wide-swath altimeters and SAR altimeters) or in the combined use of altimeter data from multiple bands. Newly reprocessed along-track measurements of Sentinel-3A, CryoSat-2, and SARAL/AltiKa altimetry missions (AVISO/TAPAS), optimized for the Arctic Ocean (retracking) have recently been produced in the framework of CNES AltiDoppler project. The tracks of the different satellite mission are then merged to provide altimetry maps with enhanced spatial coverage and resolution. This study is devoted to the exploitation of such satellite altimetry data in high-latitude regions. We investigate the benefits of the reprocessed altimetry dataset with augmented signal resolution in the context of ocean mesoscale dynamics. In particular, we perform a fit-for purpose assessment of this dataset investigating the contribution of eddy-induced anomalies to ocean dynamics and thermodynamics. This is done by co-locating eddies with Argo float profilers, in the areas representing the gateways for the Atlantic waters entering the Arctic and comparing them to fields derived from conventional altimetry maps in order to assess the added value of the enhanced altimetry reprocessing in the northern high latitude seas.

Global warming has a pronounced effect on the frequency and intensity of storm surges in the Arctic Ocean. On the one hand, changes in atmospheric conditions cause more storms to be formed in the Arctic or elsewhere that may enter the Arctic (e.g. Day & Hodges, 2018; Sepp & Jaagus, 2011). On the other hand, the Arctic Ocean is becoming increasingly exposed to atmospheric forcing due to Arctic sea ice decline (ACIA 2005, Vermaire et al. 2013). Modelling studies show that the reduced sea ice extent provides greater fetch and wave action and as such allows higher storm surges to reach the shore (Overeem et al., 2011; Lintern et al., 2011). This may cause increased erosion (e.g., Barnhart et al. 2014) and pose increased risks to fragile Arctic ecosystems in low-lying areas (e.g., Kokelj et al. 2012). In addition, Arctic surges influence global water levels, therefore the impact may also be noticeable at lower latitudes.
However, little is known about the large-scale variability in Arctic surge water levels as the data availability is compromised by environmental conditions. Long water level records from tide gauges are limited to a few locations at the coast and the high-latitudes are poorly covered by satellite altimeters. Moreover, measurements of the Arctic water level by satellite altimeters is hampered by the presence of sea ice. Here, the usage of Synthetic Aperture Radar (SAR) altimeter data provides a solution. These altimeters have a higher along-track resolution than conventional altimeters, which allows to measure water levels from fractures in the sea ice (leads) (Zygmuntowska et al., 2013). However, the location of leads changes over time, and both the temporal and spatial resolutions of the resulting water level data are highly variable. In addition, a proper removal of the tidal signal is required in order to study surge water levels. This may be particularly problematic in the Arctic as the accuracy of global tide models is reduced in polar regions (e.g. Cancet et al., 2017; Lyard et al., 2021; Stammer et al., 2014). This is can for a part be attributed to the beforementioned constrains on data availability, as well as to the seasonal modulation of Arctic tides that is not considered in most global tide models.
In the presented study we aspired to overcome the identified issues and explore the opportunities provided by SAR altimetry in studying storm surge water levels in the Arctic. For this, data are used from two high-inclination missions that are equipped with a SAR altimeter: CryoSat-2 and Sentinel-3. A classification scheme is implemented to distinguish between measurements from sea ice and leads/ocean and data stacking is applied to deal with the restricted temporal and spatial resolution. The tidal signal is removed as much as possible by applying tidal corrections from a global tide model, as well as additional corrections derived from a residual tidal analysis including seasonal modulation of the major tidal constituents. To evaluate the approach, where possible, results are compared to water levels derived from nearby tide gauges. Implications of reduced accuracy in tidal corrections are identified by analyzing the results in the light of the level of tidal activity and seasonal modulation. Finally, temporal variations in surge water levels are linked to the seasonal sea ice cycle and interannual variations in sea ice extent.

References
ASSESSMENT, ARCTIC CLIMATE IMPACT (ACIA). (2005). Impacts of a warming Arctic: Arctic climate impact assessment, scientific report
Barnhart, K. R., Overeem, I., & Anderson, R. S. (2014). The effect of changing sea ice on the physical vulnerability of Arctic coasts. The Cryosphere, 8(5), 1777-1799.
Cancet, M., Andersen, O. B., Lyard, F., Cotton, D., & Benveniste, J. (2018). Arctide2017, a high-resolution regional tidal model in the Arctic Ocean. Advances in space research, 62(6), 1324-1343.
Day, J. J., & Hodges, K. I. (2018). Growing land‐sea temperature contrast and the intensification of Arctic cyclones. Geophysical Research Letters, 45(8), 3673-3681.
Kokelj, S. V., T. C. Lantz, S. Solomon, M. F. J. Pisaric, D. Keith, P. Morse, J. R. Thienpont, J. P. Smol, and D. Esagok (2012), Utilizing multiple sources of knowledge to investigate northern environmental change: Regional ecological impacts of a storm surge in the outer Mackenzie Delta, N.W.T., Arctic, 65, 257–272.
Lintern, D. G., Macdonald, R. W., Solomon, S. M., & Jakes, H. (2013). Beaufort Sea storm and resuspension modeling. Journal of Marine Systems, 127, 14-25.
Lyard, F. H., Allain, D. J., Cancet, M., Carrère, L., & Picot, N. (2021). FES2014 global ocean tide atlas: design and performance. Ocean Science, 17(3), 615-649.
Overeem, I., R. S. Anderson, C. W. Wobus, G. D. Clow, F. E. Urban, and N. Matell (2011), Sea ice loss enhances wave action at the Arctic coast, Geophys. Res. Lett., 38, doi:10.1029/2011GL048681.
Sepp, M., and J. Jaagus (2011), Changes in the activity and tracks of Arctic cyclones, Clim. Change, 105, 577–595.
Stammer, D., Ray, R. D., Andersen, O. B., Arbic, B. K., Bosch, W., Carrère, L., ... & Yi, Y. (2014). Accuracy assessment of global barotropic ocean tide models. Reviews of Geophysics, 52(3), 243-282.
Vermaire, J. C., M. F. J. Pisaric, J. R. Thienpont, C. J. Courtney Mustaphi, S. V. Kokelj, and J. P. Smol (2013), Arctic climate warming and sea ice declines lead to increased storm surge activity, Geophys. Res. Lett., 40, 1386–1390, doi:10.1002/grl.50191.
Zygmuntowska, M., Khvorostovsky, K., Helm, V., & Sandven, S. (2013). Waveform classification of airborne synthetic aperture radar altimeter over Arctic sea ice. The Cryosphere, 7(4), 1315-1324.

It is expected that coupled air-sea data assimilation algorithms may enhance the exploitation of satellite observations whose measured brightness temperatures depend upon both the atmospheric and oceanic states, thus improving the resulting numerical forecasts. To demonstrate in practice the advantages of the fully coupled assimilation scheme, the assimilation of brightness temperatures from a forthcoming microwave sensor (the Copernicus Imaging Microwave Radiometer, CIMR) is evaluated within idealized assimilation and forecast experiments. The forecast model used here is the single-column version of a state-of-the-art Earth system model (EC-Earth), while a variational scheme, complemented with ensemble-derived background-error covariances, is adopted for the data assimilation problem.
The Copernicus Imaging Microwave Radiometer (CIMR), scheduled for the 2027+ timeframe, is a high priority mission of the Copernicus Expansion Missions Programme. Polarised (H and V) channels centered at 1.414, 6.925, 10.65, 18.7 and 36.5 GHz are included in the mission design under study. CIMR is thus designed to provide global, all-weather, mesoscale-to-submesoscale resolving observations of sea-surface temperature, sea-surface salinity and sea-ice concentration. The coupled observation operator is derived as polynomial regression from the application of the Radiative Transfer for TOVS (RTTOV) model, and we perform Observing System Simulation Experiments (OSSE) to assess the benefits of different assimilation methods and observations in the forecasts.
Results show that the strongly coupled assimilation formulation outperforms the weakly coupled one in both experiments assimilating atmospheric data and verified against oceanic observations and experiments assimilating oceanic observations verified against atmospheric observations. The sensitivity of the analysis system to the choice of the coupled background-error covariances is found significant and discussed in detail. Finally, the assimilation of microwave brightness temperature observations is compared to the assimilation of the corresponding geophysical retrievals (sea surface temperature and salinity and marine winds), in the coupled analysis system. We found that assimilating microwave brightness temperatures significantly increases the short-range forecast accuracy of the oceanic variables and near-surface wind vectors, while it is neutral for the atmospheric mass variables. This suggests that adopting radiance observation operators in oceanic and coupled applications will be beneficial for operational forecasts.

The ocean tides are one of the major contributors to the energy dissipation in the Arctic Ocean. In particular, barotropic tides are very sensitive to friction processes, and thus to the presence of sea ice in the Polar regions. However, the interaction between the tides and the ice cover (both sea ice and grounded ice) is poorly known and still not well modelled, although the friction between the ice and the water due to the tide motions is an important source of energy dissipation and has a direct impact on the ice melting. The variations of tidal elevation due to the seasonal sea-ice cover friction can reach several centimeters in semi-enclosed basins and on the Siberian continental shelf. These interactions are often simply ignored in tidal models, or considered through relatively simple combinations with the bottom friction.
In the frame of the Arktalas project funded by the European Space Agency, we have investigated this aspect with a sensitivity analysis of a regional pan-Arctic ocean tide hydrodynamic model to the friction under the sea ice cover, in order to generate more realistic simulations. Different periods of time, at the decadal scale, were considered to analyze the impact of the long-term reduction of the sea ice cover on the ocean tides in the region, and at the global scale. Tide gauge and satellite altimetry observations were specifically processed to retrieve the tidal harmonic constituents over different periods and different sea ice conditions, to assess the model simulations.
Improving the knowledge on the interaction between the tides and the sea ice cover, and thus the performance of the tidal models in the Polar regions, is of particular interest to improve the satellite altimetry observation retrievals at high latitudes, as the tidal signals remain a major contributor to the error budget of the satellite altimetry observations in the Arctic Ocean, but also to generate more realistic simulations with ocean circulation models, and thus contribute to scientific investigations on the changes in the Arctic Ocean.

The Arctic Ocean is the ocean most vulnerable to climate change. Accelerating air and ocean temperatures and deglaciation of land and sea ice alters the physical dynamics of the Arctic Ocean that impacts the sea level. Hence is sea level a bulk measure of ongoing climate related processes.

A unique feature of the Arctic Ocean is that freshwater change is the most significant contribution to sea level change. Freshwater coming from land, sea ice and rivers expands the water column and changes the dynamics of ocean currents going in and out of the Arctic Ocean. For sea level analysis of the Arctic Ocean, is the steric sea level change (change of ocean water density from temperature and salinity changes) often either inverted from satellite observations (sea surface height (SSH) from altimetry minus ocean bottom pressure (OBP) from GRACE) or based on oceanographic models that are constrained with a mix of in-situ observations, altimetry and GRACE.

Recently, studies (1,2) have shown that a satellite-independent steric sea level estimate has shown to better reconstruct the sea level features observed from altimetry compared to oceanographic models. The steric estimate (DTU Steric 2020) is composed from more than 300,000 Arctic in-situ profiles, which are interpolated into a monthly 50x50 km grid from 1990 to 2015. A further advantage is the independence of altimetry (and GRACE) and therefore ideal to be used for sea level budget analysis. Some regions with sparse in-situ observations (in particular the East Siberian Seas), showed less correlation with altimetry, but is also a region with poor tide-gauge/altimetry agreement (3,4), making it difficult to validate either of the datasets.

Here we present an update of the steric sea level product presented in (1). It now includes temperature and salinity profiles up to end 2020, representing a 31-year period from 1990-2020. Additionally, the profile data is assimilated with satellite surface salinity data from SMOS and satellite sea surface temperature data from GHRSST (Group for High-Resolution Sea Surface Temperature). Furthermore, the Arctic Ocean is divided into nine significant regions, giving a better overview of significant features and statistics of the Arctic steric sea level change. The extended timeseries allows to investigate long-term climate trends of the Arctic Ocean, which can be validated against an equal long record of altimetric sea level observations (1991-2010 up to 82N, 2011-2020 up to 88N). The dataset is useful for wide range of users looking at changes in heat content, freshwater changes, validating sea level observations (from tide gauges and altimetry) and ocean bottom pressure from GRACE/GRACE-FO (i.e. constrain leakage corrections).


1) Ludwigsen, C. A., & Andersen, O. B. (2021). Contributions to Arctic sea level from 2003 to 2015. Advances in space research, 68(2), 703-710. https://doi.org/10.1016/j.asr.2019.12.027
2) Ludwigsen, C. B., Andersen, O. B., & Kildegaard Rose, S (2021). Components of 21 years (1995-2015) of Absolute Sea Level Trends in the Arctic. Ocean Science (pre-print)
3) Armitage, T. W. K., Bacon, S., Ridout, A. L., Thomas, S. F., Aksenov, Y., & Wingham, D. J. (2016). Arctic sea surface height variability and change from satellite radar altimetry and GRACE, 2003-2014.
4) Kildegaard Rose, S., Andersen, O. B., Passaro, M., Ludwigsen, C. A., & Schwatke, C. (2019). Arctic Ocean Sea Level Record from the Complete Radar Altimetry Era: 1991-2018. Remote Sensing, 11(14), 1672. https://doi.org/10.3390/rs11141672

Recent observational and modelling studies have documented changes in the hydrography of the upper Arctic Ocean, in particular an increase of its liquid freshwater content (e.g., Haine et al. 2015, Proshutinsky et al. 2019, Solomon et al. 2021). The main factors contributing to this freshening are the melting of the Greenland ice sheet and glaciers, enhanced sea-ice melt, an increase of river discharge, increase in liquid precipitation and an increase of Pacific Ocean water influx to the Arctic Ocean through the Bering Strait. A retreating-thinning sea ice cover, and a concomitant warming-freshening upper ocean, have a widespread impact across the whole Arctic system through a large number of feedback mechanisms and interactions also with the atmospheric circulation of the northern hemisphere, having the potential to destabilize the thermo-haline circulation in the Northern Atlantic.
An increase of liquid freshwater content has been found over both the Canadian Basin and the Beaufort Sea that can have a large impact on the Arctic marine ecosystem. The importance of monitoring changes in the Arctic freshwater system and its exchange with subarctic oceans has been widely recognized by the scientific communities.
Among the key observable variables, ocean salinity is a proxy for freshwater content and allows to monitor increased freshwater from rivers or ice melt, and it sets the upper ocean stratification, which has important implications in water mass formation and heat storage. Changes in the salinity distribution may affect the water column stability and impact the freshwater pathways over the Arctic Ocean. Sea Surface Salinity (SSS) is observed from space with the L-band (1.4GHz) radiometers such as SMOS (ESA, since 2010) and SMAP (NASA, since 2015). However, retrieving SSS in cold waters is challenging, for different factors. Thanks to the ESA funded the ARCTIC+SSS ITT project, we have now a new enhanced Arctic SMOS Sea Surface Salinity product BEC v.31, which has better quality and resolution than the previous high latitude salinity products which permit to better monitor salinity changes, and thus freshwater.

In this presentation we will show the first results of the surface salinity tendency analysis done with the new SMOS BEC SSS v3.1 product in the Beaufort Sea, and other Arctic regions during summer for the period from 2011 to 2021. We will compare the results with other model (TOPAZ) and satellite observations (CryoSat and GRACE). Only summer results will be shown since, observations of SSS are feasable only when the ocean is free of ice. This preliminary analysis shows a clear freshening in the sea surface salinity the Beaufort Gyre region from 2012 to 2019.

Basal melting of floating ice shelves and iceberg calving constitute the two almost equal paths of freshwater flux (FWF) between the Antarctic ice cap and the Southern Ocean. For the Greenland ice cap the figure are quite similar even if surface melting plays a more significant role.
Basal meltwater and surface melt water are distributed over the upper few hundred meters of the coastal water column while icebergs drift and melt farther away from land.
While the northern hemisphere icebergs are, except for rare exception small (less than10km2), In the southern ocean large icebergs (larger than 100km2) act as a reservoir to transport ice far away from the Antarctic coast into the ocean interior, while fragmentation acts as a diffusive process. It generates plumes of small icebergs that melt far more efficiently than larger ones.
OGCM that include iceberg show that basal ice-shelf and iceberg melting have different effects on the ocean circulation and that icebergs induce significant changes in the modeled ocean circulation and sea-ice conditions around Antarctica or in the northern Atlantic. The transport of ice away from the coast by icebergs and the associated FWF cause these changes. These results highlight the important role plaid by icebergs and their associated FWF play in the climate system. However, there is actually no direct reliable estimate of the iceberg FWF to either validate or constrain the models.
Since 2008 the ALTIBERG project maintains a small iceberg (less than 10km2) database (North and South) using a detection method based on the analysis of satellite altimeter waveforms (http://cersat.ifremer.fr/data/products/catalogue). The archive of positions, areas, dates of icebergs as well as the monthly mean volume of ice icebergs covers now the 1992-present period.
Using classical iceberg's motion and thermodynamics equations constrained by AVISO currents, ODYSEA SST's and Wave Watch 3 wave heights, the trajectories and melting of all detected ALTIBERG icebergs are computed. The results are used to compute the daily FWF.
The FWF's temporal and spatial distribution from 1993 to 2019 are presented as well as the estimation method. The north Atlantic FWF, which has also been estimated, will also be analyzed.
Figure 1 presents the mean daily FWF, the mean daily volume of ice as well as the mean surface and thickness of the icebergs for the 1993-2019 period on a 50x50km grid.

Temperature rise and the immediate effect it has in the Arctic calls for increased monitoring of sea surface temperature (SST), which demands the highest possible synergy between the different sensors orbiting Earth, both on present and future missions. One example is the possible synergy between Sentinel-3’s SLSTR and the future Copernicus expansion satellite: Copernicus Imaging Microwave Radiometer (CIMR), which is currently in development phase. To achieve consistency between the observations from the different missions, there is a need to establish a relation between skin and subskin SSTs, which are measured by infrared and microwave sensors respectively. That will lead to the creation of more homogeneous and higher accuracy datasets that could be used to monitor climate change in greater detail and to be assimilated into climate models.

To address the aforementioned issue, the Danish Meteorological Institute (DMI) and the Technical University of Denmark (DTU) performed, on June 2021, a week-long intercomparison campaign between Denmark and Iceland, where they collected data by simultaneously deploying a microwave and an infrared radiometer side-by-side. The work was a part of the ESA funded project SHIPS4SST (ships4sst.org) and the International Sea Surface Temperature Fiducial Reference Measurement Radiometer Network (ISFRN) where shipborne IR radiometer deployments have been conducted between Denmark and Iceland for several years. In this particular campaign, two ISARs (Infrared Sea Surface Temperature Autonomous Radiometer), measuring on the 9.6 – 11.5 μm spectral band, were deployed alongside two recently refurbished EMIRADs, namely EMIRAD-C and EMIRAD-X, measuring on C and X band respectively.

This study aims at demonstrating the methodology applied within ESA CCI SST to retrieve SST from the microwave brightness temperature, and present a first attempt to establish a relationship between skin and subskin SST, as well as the overall research progress so far.

Whilst the Arctic Ocean is relatively small, containing only 1% of total ocean volume, it receives 10% of global river runoff. This river runoff is a key component of the Arctic hydrological cycle, providing significant freshwater exchange between land and the ocean. Of this runoff, Russian rivers alone contribute around half of the total river discharge, or a quarter of the total freshwater to the Arctic Ocean, predominantly to the Kara and Laptev Seas. In these seas, inflowing riverine freshwater remains at the surface, helping to form the cold, fresh layer that sits above inflowing warm and salty Atlantic Water; this sets up the halocline that governs Eurasian shelf sea, and wider Arctic Ocean, stratification. This fresh surface layer prevents heat exchange between the underlying Atlantic Water and the overlying sea ice, limiting sea ice melt and strengthening the existing sea ice barrier to atmosphere-ocean momentum transfer. However, the processes that govern variability in riverine freshwater runoff and its interactions with sea ice are poorly understood and are key to predicting the future state of the Arctic Ocean. Understanding these processes is particularly important in the Laptev Sea as a source region of the Transpolar Drift and a key region of sea ice production and deep water formation (Reimnitz et al., 1994).

Over most of the globe, L-band satellite acquisitions of sea surface salinity (SSS), such as from Aquarius (2011–2015), SMOS (2010- present), and SMAP (2015-present), provide a new tool to study freshwater storage and transport. However, the low sensitivity of L-band signal in cold water and the presence of sea ice makes retrievals at high latitudes a challenge. Nevertheless, retreating Arctic sea ice cover and continuous progress in satellite product development make the satellite based SSS measurements of great value in the Arctic. This is particularly evident in the Laptev Sea, where gradients in SSS are strong and in situ measurements are sparse. Previous work has demonstrated a good consistency of satellite based SSS data against in situ measurements, enabling greater confidence in acquisitions and making satellite SSS data a truly viable potential in the Arctic (Fournier et al., 2019; Supply et al., 2020).

This study combines satellite based SSS data, in-situ observations and reanalysis products to study the roles of Lena river discharge, ocean circulation, vertical mixing and sea ice cover on interannual variability in Laptev Sea dynamics. Comparison of two SMOS products, SMOS LOCEAN CEC L3 Arctic v1.1 and SMOS BEC Arctic L3 v3.1, and two SMAP SSS products, SMAP JPL L3 v5.04.2 and SMAP RSS L3 v4.03 were considered. Whilst the general patterns of salinity are broadly similar in all products, their patterns differ interannualy, with particular discrepancies in magnitude. Interannual variability in the LOCEAN SMOS SSS closely resembles that in both SMAP products, most notably in the magnitude and direction of Lena river plume propagation. However, the mean state of SMAP RSS SSS is much fresher than the other products. Comparison against TOPAZ reanalysis highlights similar interannual pattern with both SMAP products and SMOS LOCEAN, but with lower amplitude. The close resemblance of the SMOS CEC LOCEAN and the SMAP products gives confidence in using the full SMOS LOCEAN timeseries (2012-present) to study interannual variability on a 10-year time scale. The full SMOS LOCEAN timeseries shows that two years (2018 and 2019) stand out as having a much larger, fresher river plume than other years. However, the larger plume in these years doesn’t appear to be caused by increases in Lena river runoff. Numerical model output, in-situ data and satellite products are used to study the cause of this variability.


 
Bibliography
Fournier, S., Lee, T., Tang, W., Steele, M., Olmedo, E., 2019. Evaluation and Intercomparison of SMOS, Aquarius, and SMAP Sea Surface Salinity Products in the Arctic Ocean. Remote Sens. 11, 3043. https://doi.org/10.3390/rs11243043
Reimnitz, E., Dethleff, D., Nürnberg, D., 1994. Contrasts in Arctic shelf sea-ice regimes and some implications: Beaufort Sea versus Laptev Sea. Mar. Geol., 4th International Conference on Paleoceanography (ICP IV) 119, 215–225. https://doi.org/10.1016/0025-3227(94)90182-1
Supply, A., Boutin, J., Vergely, J.-L., Kolodziejczyk, N., Reverdin, G., Reul, N., Tarasenko, A., 2020. New insights into SMOS sea surface salinity retrievals in the Arctic Ocean. Remote Sens. Environ. 249, 112027. https://doi.org/10.1016/j.rse.2020.112027

Sea level observations from satellite altimetry in the Arctic Ocean are severely limited due to the presence of sea ice. To determine sea surface heights and enable studies of the ocean surface circulation, it is necessary to first detect openings in the sea ice cover (leads and polynyas) where the ocean surface is exposed. This is of particular interest in the coastal areas of the Arctic, where glaciers calve into the Arctic Ocean. The increasing freshwater influx in the last years leads to changes in the sea level and the thermohaline circulation.

The ESA Explorer mission Cryosat-2 was launched in 2010, aiming at the monitoring of the cryosphere. The satellite works in three different acquisition modes. One of these modes is the interferometric SAR (InSAR) mode. The radar returns (called waveforms) of this mode are characterized by a higher temporal resolution, which allows a more reliable detection of leads and polynyas in coastal areas. An unsupervised classification approach based on Machine Learning is implemented for Cryosat-2 InSAR waveforms. The classification approach utilizes differences in scattering properties from sea ice, open ocean, and calm enclosed ocean. By defining quantitative parameters from the waveform shape, the waveforms are grouped by comparing the similarity of the parameters without the necessity of pre-classified data. The classification performance is validated against optical images of spatiotemporally overlapping aircraft overflights. An algorithm is implemented to automatically detect leads from the optical images while minimizing the time difference between altimetry and optical observations.

The implementation of an unsupervised detection of open water in the Arctic Ocean environment is part of the recently launched AROCCIE project (ARctic Ocean surface circulation in a Changing Climate and its possible Impacts on Europe). The aim of the project is to combine satellite altimetry with numerical ocean modelling to determine changes in Arctic Ocean surface circulation from 1995 to present. AROCCIE will use the classification of InSAR data to create a more comprehensive dataset of sea surface heights for further analysis of ocean circulation changes in the vicinity of the Arctic's rugged coastlines.

Accurate sea surface temperature (SST) observations are crucial for climate monitoring, understanding of air-sea interactions as well as in weather and sea ice forecasts through assimilation into ocean and atmospheric models. In general, two types of retrieval algorithms have been used to retrieve SST from passive microwave satellite observations: statistical algorithms and physical algorithms based on the inversion of a radiative transfer model (RTM). The physical algorithms are constrained by the accuracy of the RTM and the representativeness of the of the observation and prior error covariances. They can be used to identify measurement errors but require ad-hoc corrections of the geophysical retrievals to take these into account. Statistical algorithms may account for some of the measurement errors through the coefficient derivation process, but the retrievals are limited to the established relationships between the input variables. Machine learning (ML) algorithms may supplement or improve the existing retrieval algorithms through their higher flexibility and ability to recognize complex patterns in data.
In this study, several types of ML algorithms have been trained and tested on the global ESA SST CCI multi-sensor matchup dataset, with a focus on their performances in the Arctic region. The machine learning algorithms include two multilayer perceptron neural networks (NNs) and different types of ensemble algorithms e.g. a random forest algorithm and two boosting algorithms: least-squares boosting and the Extreme Gradient Boosting (XGB). The algorithms have been evaluated for their capability to retrieve SST from passive microwave (PMW) satellite observed brightness temperatures from the Advanced Microwave Scanning Radiometer – Earth Observing System (AMSR-E). To validate the algorithms independent SST observations from drifting buoys have been used. The performance of the ML algorithms has been compared and evaluated against the performance of an existing state of the art regression (RE) algorithm with a focus on the Arctic. In general, the ML algorithms show good global performances with decreasing performances towards higher latitudes. The XGB algorithm performs best in terms of bias and standard deviation followed by the NNs and the RE algorithm. The boosting algorithms and the NNs are able to reduce the bias in the Arctic compared to the other ML algorithms. For each of the ML algorithms, the sensitivity (i.e. the change in retrieved SST per unit change in the true SST) has been estimated for each matchup by using simulated brightness temperatures from the Wentz/DMI forward model. In general, the sensitivities are lower in the Arctic compared to the global averages. The highest sensitivities are found using the neural networks, and the lowest using the XGB algorithm, which underlines the importance of including sensitivity estimates when evaluating retrieval performances.
The good performance of the ML algorithms compared to the state of the art RE algorithm in this initial study demonstrates that there is a large potential in the use of ML techniques to retrieve SST from PMW observations. The ML methodology, where the algorithms select the important features based on the information in the training data, work well in complex problems where not all physical and/or instrumental effects are well determined. A suitable ML application could e.g. be in a commissioning phase of new satellites (e.g. for the Copernicus Imaging Microwave Radiometer (CIMR) developed by ESA).

The Arctic has warmed more than twice the global rate, which makes it a crucial region to monitor surface temperatures in this region. Global surface temperature products are fundamental for assessing temperature changes, but for the Arctic sea ice these products are traditionally only built on near-surface air temperature measurements from weather stations and sparse drifting buoy temperature measurements. However, only limited in situ observations are actually available in the Arctic due to the extreme weather conditions and limited access. Therefore, satellite observations have a large potential to improve the surface temperature estimates in the Arctic Ocean due to the good temporal and spatial coverage.
We present the first satellite derived combined and gap-free (L4) climate data set of sea surface temperatures (SST) and ice surface temperatures (IST) covering the Arctic Ocean (>58°N) for the period 1982-2021. The derived L4 SST/IST climate data set has been generated as a part of the Copernicus Marine Environment Monitoring Service (CMEMS) and the National Centre for Climate Research (NCKF). The data set has been generated by combining multi-satellite observations using statistical optimal interpolation (OI) to obtain daily gap-free fields with a spatial resolution of 0.05°. Due to the different characteristics of the open ocean, sea ice and the marginal ice zone (MIZ), the OI statistical parameters have been derived separately for each region. Therefore, it is very important with an accurate sea ice concentration (SIC) field for identifying the regions. Here, a combination of several SIC products and additional filtering have been used to produce an improved SIC product.
Observations from drifting buoys, moored buoys, ships and the Icebridge campaigns have been used to validate the L4 SST/IST over the ocean and sea ice. The combined sea and ice surface temperature of the Arctic Ocean provides a consistent climate indicator which can be used to monitor day-to-day variations as well as long term climate trends. The combined sea and ice surface temperature of the Arctic Ocean has increased with more than 4°C over the period from 1982 to 2021.

Like other areas of climate science, Arctic Sea ice forecasting can be improved by using advanced data assimilation (DA) to combine model simulations and observations. We consider the Ensemble Kalman filter (EnKF), one of the most popular DA methods, widely used in climate modelling systems. In particular, we apply a deterministic Ensemble Kalman filter (DEnKF) to the Lagrangian sea ice model neXtSIM (neXt-generation Sea Ice Model). neXtSIM implements a novel Brittle Bingham–Maxwell sea ice rheology, computationally solved on a time-dependent evolving mesh. This latter aspect represents a key challenge to the EnKF as ensemble member size and nodes are generally different. The DEnKF analysis is performed on a fixed reference mesh via interpolation and then projected back to the individual ensemble meshes. We propose an ensemble-DA forecasting system for Arctic sea ice forecast by assimilating the OSI-SAF sea ice concentration (SIC) and the CS2SMOS sea ice thickness (SIT). The ensemble is generated by perturbing atmospheric and oceanic forcing online throughout the forecast. We evaluate the impact of sea-ice assimilation on the Arctic winter sea-ice forecast skills against the satellite observations and a free run during the 2019-2020 Arctic winter. We have obtained significant improvements in SIT but fewer improvements regarding the other ice states. The improvements are mainly due to assimilating the relevant observations. It also shows that neXtSIM as a stand-alone sea ice model is computationally efficient by using external forcing which has assimilated observations and keep good forecast skills.

Traditional global sea surface temperature (SST) analyses have often struggled in high-latitude regions. The challenges that exist are numerous (sea-ice cover, cloud, perpetual dark, perpetual sunlight, insufficient in situ for validation & bias correction, anomalous atmospheric conditions). In this presentation, we outline the prospects for a new high-resolution sea surface temperature analysis specifically for the Arctic region. There are many reasons why such a product is desirable now. Firstly, the Arctic region is anticipated to be the most sensitive to climatic change, and has already experienced a number of substantially anomalous years. Sea ice cover has been decreasing, and yet is still highly variable. The development and progression of the polar front has a major influence on mid-latitude Northern Hemisphere weather patterns (storm tracks, cold air outbreaks, etc.). Accurate knowledge of high-latitude sea surface temperature is crucial for the prediction of sea ice growth and decay, along with estimation of air-sea fluxes, ecological processes and monitoring of overall conditions. Many research areas within the Arctic section of this symposium will benefit from such a dataset.

An equally important aspect of this presentation is the illustration of limitations in existing SST products in the Arctic region. This is particularly important for end-users who may be utilizing products while being largely unaware of the issues. The biggest challenge is ensuring that the available data are fully exploited, i.e. that potentially valid observations are excluded due to quality control (cloud screening, etc.) procedures that have not been optimized for the Arctic region. We use matches with high-latitude saildrone data to explore the impact of current cloud detection schemes and indicate how improvements can be made. Similarly, ice masking may deprive users of valuable observations in the marginal ice zone. Other issues we explore include the correction for atmospheric effects in Arctic atmospheres, which are out-of-family compared with lower latitude oceans where algorithms have been developed and validated. In this regard, we show that the dual-view capabilities of the Sentinel-3 SLSTR instrument can provide a valuable reference. The need for significantly different approaches to quality control and assimilation are explained, along with the need for proxy observations under sea ice. The interdependence of these observation types and models requires a coordinated approach in order to achieve success.

Ocean surface current is yet poorly observed by satellite remote sensing in comparison to other sea surface variables, such as the surface temperature, the surface winds and waves field, among others. Over the recent years, a rich of radar missions dedicated to the ocean current mapping at the global scale have emerged. Most of the proposed techniques take advantage of the Doppler shift obtained by the phase-resolving radars, which is associated with the sea surface motion. Both observational and simulation efforts have demonstrated that apart from the satellite motion, the total Doppler shift is composed of contributions from the surface winds, ocean waves, in addition to the underlying ocean surface current. The knowledge of concurrent wind and wave component is essential for the removal of their impact to acquire the geophysical Doppler shift. Such Doppler shift residual could be directly converted to the line-of-sight current velocity. Successful applications of this technique to observe major ocean current have been illustrated based on the single-antenna synthetic aperture radar systems, which proves the feasibility of Doppler shift method for further exploitation. A radar mission designed for concurrent measurements of wind, waves and current is under development. As the preparatory studies for this mission, we conduct the simulation of Doppler shift from sea surface with a wide variety of sea state conditions and different radar configurations (polarization, radar frequency and incidence angles et al.). Results illustrate that the contribution of winds and waves constitutes a major part of total Doppler shift, particularly when the underlying surface current is relatively weak. This further evidences the necessity to remove the wind/wave component for accurate retrieval of surface current in the future operational processing. Given the variable sensitivity of polarization to the ocean waves, dual-polarized Doppler shift brings more information than single-polarized channel, which could be promoted in the radar system configuration. The simulation study strengthens our confidence on this pending mission to enhance the observational capability of ocean surface current on top of other concepts.

Using signals of opportunity (SoOP), i.e. signals already transmitted for uses different from remote sensing, is an advantageous way to carry out bi-static observations at a reduced cost, as the transmitter is already operated for its primary use. Consolidated examples of this are GNSS Radio Occultation measurements [e.g., 1] and GNSS Reflectometry measurements [e.g., 2, 3] done from space.

Different research projects have been carried out during the last decade in order to use SoOP at higher frequencies e.g. [4, 5], and thus shorter wavelengths, to study the ocean surface. Parameters of interest are the sea surface roughness and sea surface altimetry. Candidate source of opportunity are FM-radio or digital satellite TV signals broadcasted from geostationary orbit. In particular, digital satellite TV signals have a very large potential thanks to i.) the large number of broadcasting satellites (~300), ii.) their extremely large total bandwidth which can span up to 2 GHz when many TV channels are considered, and iii.) the stronger available power compared to GNSS signals. This results in an expected precision of few cm in altimetric sea-surface observations [6].

In addition to these potentialities, when considering the larger available power, digital satellite TV signals can be used in bi-static geometries different from forward scattering. In these geometries, the Doppler signature of the reflected signals is also affected by the horizontal movement of the reflecting target. Thus, the horizontal velocity component of the ocean waves and the ocean current will affect the Doppler frequency of the reflected signal. In addition, as the wavelength of these signals is shorter (λ~2.5 cm) the Doppler frequency will have also a larger value compared to GNSS signals. An experimental demonstration of estimating the water velocity of a river using digital satellite TV signals can be found in [7].

In August 2021, an experimental campaign was carried out at the Majorca Island. On top of its highest peak (Puig Major 1480 m) two antennas were installed. The first one was used to acquire the direct TV signals transmitted from the ASTRA 1M (19.2E) satellite. The second antenna was pointed towards the sea to collect the signals that bounced off the sea surface, in non-specular but back- and side-scattering geometry. Direct and reflected signals were down-converted to IF, digitized at 80 Msps and stored on a SSD hard drive. Different data acquisitions were carried out in a variety of conditions: Signals of different TV channels were used, thus having diversity in wavelength. The down-looking antenna was also pointed at different elevation angles and azimuths with respect to the direction of the waves/currents.

The recorded data are being post processed. We will present preliminary results of the experimental campaign in order to try to establish the main aspects to be considered in a future airborne or space-borne instrument for a cost-effective direct measurement of the sea surface currents.


[1] Kursinski, E. R., Hajj, G. A., Schofield, J. T., Linfield, R. P., & Hardy, K. R. (1997). Observing Earth's atmosphere with radio occultation measurements using the Global Positioning System. Journal of Geophysical Research: Atmospheres, 102(D19), 23429-23465.


[2] Foti, G., Gommenginger, C., Jales, P., Unwin, M., Shaw, A., Robertson, C., & Rosello, J. (2015). Spaceborne GNSS reflectometry for ocean winds: First results from the UK TechDemoSat‐1 mission. Geophysical Research Letters, 42(13), 5435-5441

[3] Ruf, C. S., Atlas, R., Chang, P. S., Clarizia, M. P., Garrison, J. L., Gleason, S., ... & Zavorotny, V. U. (2016). New ocean winds satellite mission to probe hurricanes and tropical convection. Bulletin of the American Meteorological Society, 97(3), 385-395.

[4] Ribó, S., Arco, J. C., Oliveras, S., Cardellach, E., Rius, A., & Buck, C. (2014). Experimental results of an X-Band PARIS receiver using digital satellite TV opportunity signals scattered on the sea surface. IEEE Transactions on Geoscience and Remote Sensing, 52(9), 5704-5711.

[5] Shah, R., Garrison, J. L., & Grant, M. S. (2011). Demonstration of bistatic radar for ocean remote sensing using communication satellite signals. IEEE Geoscience and Remote Sensing Letters, 9(4), 619-623.

[6] Shah, R., Garrison, J., Ho, S. C., Mohammed, P. N., Piepmeier, J. R., Schoenwald, A., ... & Bradley, D. (2017, July). Ocean altimetry using wideband signals of opportunity. In 2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) (pp. 2690-2693). IEEE.

[7] Ribó, S., Cardellach, E., Fabra, F., Li, W., Moreno, V., & Rius, A. (2018, September). Detection and Measurement of Moving Targets Using X-band Digital Satellite TV Signals. In 2018 International Conference on Electromagnetics in Advanced Applications (ICEAA) (pp. 224-227). IEEE.

We undertake numerical experiments to show how observations of the geostrophic currents based on satellite data like the Sentinel-1 RVL products would influence and potentially improve the ”geodetic” (i.e. satellite-based only) estimation of the mean dynamic topography. The dynamic topography is the divergence of the sea surface from a hypothetical ocean at rest (the geoid) resulting from various “dynamic” processes. In particular, the mean dynamic topography is related to the steady state circulation in the oceans and consequently has meaning for studying global mass and heat transport. In this study we restrict ourselves to a mean model of the dynamic topography and assume a static gravity field. A purely observation driven approach is the joint estimation by means of a least-squares adjustment in which the sea surface height as measured by satellite altimetry is modelled as the sum of the geoid undulation and the dynamic topography. Supplementary to altimetric observations are gravity field solutions obtained from space missions, e.g. GRACE and/or GOCE, that are required to separate the two signals. Such an approach yields a so-called geodetic model of the dynamic topography that is independent of strictly oceanographic models that implement ocean physics. This enables its use in validation of oceanographic models as well as providing input data for combined models (“data assimilation”). A great challenge of the geodetic approach lies in the inconsistencies in spatial resolution between the different observation types. While the altimetry data boasts high resolution along-track (across-track depends on mission), the gravity field data is coarser on the order of one or two magnitudes. Thus it is difficult to separate the higher frequency signal that can be seen in the altimetry. For this to succeed it is required to introduce either higher resolution gravity data and/or a sufficiently accurate and preferably homogeneously sampling source of information for the dynamic topography, both under the premise of being satellite-only. Our hypothesis is that a huge opportunity comes with Doppler-derived surface current velocity measurements from SAR-satellites like Sentinel-1. Assuming the feasibility of reducing these observations to reflect geostrophic surface currents, these can be directly mapped to the spatial gradient of the dynamic topography. Such data points then provide exclusive information in the joint estimation, yielding a more stable separation. The presented study evaluates the potential gains that could be achieved by incorporating satellite-based measurements of the geostrophic surface currents, e.g. reduced Sentinel-1 RVL WV-mode type observations.

Spaceborne doppler scatterometer is a newly developed radar for ocean surface wind and current field remote sensing. Direct measurement of the global ocean surface current is of great scientific interest and application value for understanding multiscale ocean dynamics, air-sea interaction, ocean mass and energy balance, and ocean carbon budget as well as their variabilities under climate change. DOPpler Scatterometer (DOPS) onboard Ocean Surface Current multiscale Observation Mission(OSCOM) is a dual frequency doppler radar can directly measure ocean surface currents with a high horizontal resolution of 5–10 km and a swath larger than 1000km. DOPS is proposed to use a real-aperture radar with a conically scanning system. The designed satellite orbit is a sun-synchronous orbit with an altitude of 600 km and a field angle from 46° - 48°, corresponding to a ground swath larger than 1000 km. The geometry of the Doppler Scatterometer observationis shown in Figure 1. The aperture of the antenna is about 2 meters, so the beam width of DOPS is about 0.3deg in Ka band and 0.8deg in Ku band. Thus，the azimuth resolution is better than 5 km in Ka band and 10km in Ku band respectively at an orbital altitude of 600km.
At system level, an end to end simulation of DOPS is carried out to evaluate the doppler accuracy. The system noise and nonlinear effect, such as the distortion caused by power amplifier are simulated to evaluated the doppler detection errors. Based on these simulations, the transmit power of DOPS is evaluated to ensure the enough echo SNR for doppler detection. For power amplifier simulations, both Rapp nonlinear model (for solid state amplifiers) and Saleh nonlinear model (for solid state amplifiers) are established. For noise simulations, random noise and phase noise are injected to sea surface echoes. As a result, nonlinear distortion of power amplifier will cause about 0.01m/s doppler error at a low saturation status. For a wide band random noise, 10dB SNR will cause about 0.02-0.03m/s doppler error. The doppler measurement errors of incidence angle and observation azimuth are also evaluated. These errors are caused by satellite attitude determinations, where satellite attitude contains the pitch, the yaw, and the roll. As the results of simulation, to achieve the current velocity accuracy better than 0.1m/s, the measurement errors of incidence angle should be smaller than 0.001°, and satellite velocity should be smaller than 0.01 m/s.

Ku-band observations from scatterometers are easier affected by rain due to their shorter wavelength than those collected at C-band, while both frequencies are commonly applied for wind scatterometers. We proposed a support vector machine (SVM) model based on the analysis of Quality Control (QC) indicators of rain screening ability, which has been validated by collocated winds from the Ku-band and C-band scatterometers: OSCAT-2 and ASCAT-B onboard the ScatSat and MetOp-B satellites respectively, together with simultaneous rain rates from the Global Precipitation Mission (GPM) products. Meanwhile, the principle of SVM was addressed for its advantages in the rain-effect correction problem. The established SVM model was evaluated by the testing set not applied in the training procedure. In the verification, where QC-accepted winds from the C-band collocations that are of QC accepted features are applied as the truth, given their low rain sensitivity.

In this research, first, the data sets are increased by including the collocations from the OCSAT-2 and the ASCAT-A scatterometers. The wind speed range applied for the model has been extended based on the recent update of the QC indicator, Joss, which is one of the inputs for the SVM. Then to validate the model, the probability density functions (pdf) and the features due to rains of the inputs due to rain are checked in more detail. The results of the SVM from the new test set, which is not applied in the training procedure, are analyzed specifically, in addition to the statistical features obtained comparing the resulting winds and the truth. Along with the pdf, cumulative density functions (CDF) are also checked. A case study is conducted with simultaneous references from the Medium Infrared advanced imager on board the Himawari-8 satellite.

We conclude that the corrected winds can provide improved quality information for Ku-band scatterometers under rain that can be vital for nowcasting applications, where the effectiveness of optimization methods based on Machine Learning for such problems is proven.
In this research, we also discuss the application of joint-SVMs for better representing wind-rain tangling problem and the possibility of resolving winds and rain rates in such model.

Australia has a vast marine estate and amongst the longest coastlines in the World. Offshore ocean wind measurements are necessary for monitoring for a variety of users such as offshore industries (oil and gas, fisheries etc.) and understanding wind climatology for offshore operations, ship navigation, and coastal management. Australia also has a developing offshore wind energy industry. However, there are few sustained in-situ coastal ocean surface wind measurements around Australia, and either remain largely limited to reefs, jetties and coastal infrastructure or are acquired commercially by offshore industry operators. One exception is the ocean wind record from Southern Ocean Frequency Series (SOFS) flux station (Schulz et al., 2012) several hundred kms offshore south-west of Tasmania.

Sentinel-1 A and B Synthetic Aperture Radar (SAR) satellites regularly map the wider Australian coastal region and provide an opportunity to exploit these data to compile an up-to-date database of coastal wind measurements. Such a high-resolution coastal winds database from SAR also compliments global Scatterometer wind measurements as Scatterometers provide limited data closer to the shore. Two such valuable SAR winds databases already exist in other geographical regions, including NOAA’s operational SAR derived wind products (Monaldo et al., 2016) primarily focused on North America and DTU (Technical University of Denmark) Wind Energy’s SAR winds database (Hasager et al., 2006) with a European focus. With this goal in sight, a regional calibrated coastal SAR winds database has been developed for the Australian region from Sentinel-1 missions.

SAR winds are derived using input data from Sentinel-1 level-2 ocean winds (owi) product (CLS, 2020) sourced from the Copernicus Australasia regional data hub. The owi product contains all the input variables necessary to derive SAR winds including normalised radar cross section (NRCS), local incidence angle, satellite heading, and collocated model wind speed and direction from ECMWF. The algorithm applied for wind inversion is based on a variational Bayesian inversion approach as proposed in Portabella et al. 2002, and the Sentinel-1 Ocean wind algorithm definition document (CLS, 2019) with CMOD5.N as the underlying geophysical model function – GMF (Hersbach et al. 2010). For consistency, the whole Sentinel-1 archive is processed using the same wind inversion scheme and GMF. The resulting spatial resolution of the derived winds is roughly 1 km, like the owi product. The winds are also quality flagged in a systematic manner by using the ratio of the measured to simulated NRCS as a proxy for quality of the winds retrieved and applying various thresholds of median absolute deviation to this ratio. As in-situ measurements are not available in the region, calibration of SAR wind speed is performed against the calibrated (against NDBC buoy wind speeds) Metop-A and B Scatterometer winds database (Ribal et al., 2020) matchups. Calibrated SAR wind speeds are then validated against an independent Altimeter wind speed database (Ribal et al., 2019).

Such a high-resolution coastal winds archive has numerous uses for various applications. The intention is to explore these data in the future for suitability in offshore wind resource assessment, better understanding of coastal wind climatology alongside other regional model hindcast and reanalyses data, and verification of model wind fields, whose quality is a major source of error in wave models.

References

Hasager, C.B., Barthelmie, R.J., Christiansen, M.B., Nielsen, M. and Pryor, S.C. (2006), Quantifying offshore wind resources from satellite wind maps: study area the North Sea. Wind Energ., 9: 63-74. https://doi.org/10.1002/we.190

Hersbach, H. (2010). Comparison of C-Band Scatterometer CMOD5.N Equivalent Neutral Winds with ECMWF, Journal of Atmospheric and Oceanic Technology, 27(4), 721-736.

Monaldo, F. M., Jackson, C. R., Li, X.; Pichel, W. G. Sapper, J., Hatteberg, R. (2016). NOAA high resolution sea surface winds data from Synthetic Aperture Radar (SAR) on the Sentinel-1 satellites. NOAA National Centers for Environmental Information. Dataset. https://doi.org/10.7289/v54q7s2n

Portabella, M., Stoffelen, A., and Johannessen, J. A., (2002). Toward an optimal inversion method for synthetic aperture radar wind retrieval, J. Geophys. Res., 107(C8), doi:10.1029/2001JC000925.

Ribal, A., Young, I.R. (2019). 33 years of globally calibrated wave height and wind speed data based on altimeter observations. Sci Data 6, 77. https://doi.org/10.1038/s41597-019-0083-9

Ribal, A., & Young, I. R. (2020). Calibration and Cross Validation of Global Ocean Wind Speed Based on Scatterometer Observations, Journal of Atmospheric and Oceanic Technology, 37(2), 279-297. https://doi.org/10.1175/JTECH-D-19-0119.1

Schulz, E. W., Josey, S. A., & Verein, R. (2012). First air-sea flux mooring measurements in the Southern Ocean. Geophysical Research Letters, 39(16). http://dx.doi.org/10.1029/2012GL052290

Sentinel-1 Ocean Wind Fields (OWI) Algorithm Theoretical Basis Document (ATBD). (2019). Collecte Localisation Satellites (CLS). Ref: S1-TN-CLS-52-9049 Issue 2.0. Jun 2009.

Sentinel-1 Product Specification. (2020). Collecte Localisation Satellites (CLS). Ref: S1-RS-MDA-52-7441. Issue 3.7. Feb 2020.

EUMETSAT, the European Organisation for Meteorological Satellites, is expanding its scope beyond supporting meteorology, environment and climate monitoring on a global scale, to oceanography. To this end, EUMETSAT operates satellites and data processing systems, including Ocean and Sea Ice Satellite Application Facilities, to provide services which are of high value to ocean monitoring and prediction.
Current EUMETSAT programmes, as well as the European Copernicus programme of which EUMETSAT is a delegated entity, provide operational observations of sea and sea ice. The EUMETSAT marine portfolio includes surface temperature, ocean vector winds, sea surface topography, sea ice parameters, ocean colour and other key marine products.
We will review recent innovations in the EUMETSAT stream of marine satellite data, from the Sentinel-3 constellation, the Sentinel-6 Michael Freilich mission and the EPS/ASCAT mission. Upcoming and planned evolutions responding to the needs of ocean monitoring and prediction users will be presented.

Ocean surface wind vector is of paramount importance in a broad range of applications including wave forecasting, weather forecasting, and storm surge [R1-R5].
The primary remote sensing instruments for wind field retrieval from space is the microwave scatterometer. Although the latter calls for a spatial sampling adequate for several climatological and meso-scale applications, severe limitations to the use of scatterometer products arise when dealing with regional-scale applications. In contrast, the Synthetic Aperture Radar (SAR) achieves a finer spatial resolution and therefore has the potential to provide wind field information with much more spatial details. This can be important in several applications, such as in semi enclosed seas, in straits, along marginal ice zones, and in coastal regions, where scatterometer measurements are contaminated by backscatter from land and ice and the wind vector fields are often recognized to be highly variable. In such regions, wind field estimates retrieved from SAR images would be very desirable.
In this study, the main outcomes related to the Italian Space Agency (ASI) funded project APPLICAVEMARS, whose goal is estimating the ocean surface wind vector using L-, C- and X-band SAR imagery, are presented. The wind processor developed to estimate sea surface wind field from L-band SAOCOM, C-band Sentinel-1A/B and X-band CSK/CSG SAR imagery is described through some thought showcases where:
a) the scatterometer-based Geophysical Model Function is forced using both external (SCAT/ECMWF) and SAR-based wind directions, the latter evaluated by the developed methodologies based on the 2D Continuous Wavelet Transform [6] and Convolutional Neural Network [7] at high spatial resolution (1 km);
b) the wind field is estimated over collocated L-, C- and X-band SAR imagery to study both the aspects related to the GMFs and those dependent on the capacity of the different SAR frequencies to reveal the wind spatial structures.

[R1] Chelton D. B., M. G. Schlax, M. H. Freilich, R. F. Milliff, 2004: Satellite measurements reveal persistent small-scale features in ocean winds. Science, 303, 978- 983, doi:10.1126/science.1091901.
[R2] Lagerloef, G., R. Lukas, F. Bonjean, J. Gunn, G. Mitchum, M. Bourassa, and T. Busalacchi, 2003: El Niño tropical Pacific Ocean surface current and temperature evolution in 2002 and outlook for early 2003. Geophys. Res. Lett., 30, 1514, doi:10.1029/2003GL017096.
[R3] Gierach, M. M. M. A. Bourassa, P. Cunningham, J. J. O'Brien, and P. D. Reasor, 2007: Vorticity-based detection of tropical cyclogenesis. J. Appl. Meteor. Climatol., 46, 1214-1229, doi:10.1175/JAM2522.1.
[R4] Isaksen, L., A. Stoffelen, 2000: ERS-Scatterometer wind data impact on ECMWF's tropical cyclone forecasts. IEEE Trans. Geosci. Rem. Sens., 38, 1885-1892.
[R5] Morey, S. L., S. R. Baig, M. A. Bourassa, D. S. Dukhovskoy, and J. J. O'Brien, 2006: Remote forcing contribution to storm-induced sea level rise during Hurricane Dennis, Geophys. Res. Lett., 33, L19603, doi:10.1029/2006GL027021.
[6] Zecchetto, S., Wind Direction Extraction from SAR in Coastal Areas, Remote Sensing,10(2), 261, 2018 (doi:10.3390/rs10020261)
[7] Zanchetta, A. and S. Zecchetto, Wind direction retrieval from Sentinel-1 SAR images using ResNet, Remote Sensing of Environment, 253, 2021 (https://doi.org/10.1016/j.rse.2020.112178)

Microwave scatterometers play a key role when dealing with operational surface wind measurements. However, their relatively coarse spatial resolution triggered the development of wind retrieval techniques based on synthetic aperture radar (SAR) measurements. Commonly used techniques are based on the normalized radar cross section (NRCS) or radar backscatter and several empirical geophysical model functions (GMFs), originally developed to exploit C-band VV-polarized scatterometer measurements, have been tuned and recalibrated to deal with SAR measurements at different frequencies and polarizations [1]-[3]. The radar backscatter is sensitive to both wind speed and wind direction; hence, the latter must be available to constrain the GMFs [4] when retrieving the wind speed. Such a technique is limited by the fact that errors in the wind direction estimation are propagated into the wind speed estimation [5].
The so-called azimuth cutoff technique, originally proposed by Kerbaol et al. [6] to derive significant wave height (SWH) and sea surface wind speed, does not need neither calibration of the data nor any a priori information on wind direction and, therefore, has recently gained more attention [7].
In this study sea surface wind estimation is addressed using both scatterometer-based GMFs and the azimuth cut-off technique using a data set of Sentinel-1A/B SAR imagery where collocated HY2-A scatterometer wind estimates (on a 25km spatial grid) are available. The proposed rationale aims at proving that SAR NRCS, averaged on a 25km grid, is consistent with the HY2-A NRCS. Two steps will be accomplished: 1) estimating sea surface wind speed using SAR imagery through the scatterometer-based GMFs forced by the HY2-A wind direction and contrasting it with the HY2-A wind speed and with estimates obtained using the azimuth cut-off; 2) estimating the wind direction from the scatterometer-based GMF forced by the azimuth cut-off wind speed and contrasting it with the HY2-A wind direction;

[1] A. A. Mouche et al., “On the use of Doppler shift for sea surface wind retrieval from SAR,” IEEE Trans. Geosci. Remote Sens., vol. 50, no. 7, pp. 2901–2909, Jul. 2017.
[2] G. Grieco, F. Nirchio, and M. Migliaccio, “Application of state-of-the- art SAR X-band geophysical model functions (GMFs) for sea surface wind (SSW) speed retrieval to a data set of the Italian satellite mission COSMO-SkyMed,” Int. J. Remote Sens., vol. 36, no. 9, pp. 2296–2312, 2015.
[3] Y. Ren, S. Lehner, S. Brusch, X. Li, and M. He, “An algorithm for the retrieval of sea surface wind fields using X-band TerraSAR-X data,” Int. J. Remote Sens., vol. 33, no. 23, pp. 7310–7336, 2012.
[4] C. C. Wackerman, C. L. Rufenach, R. A. Shuchman, J. A. Johannessen, and K. L. Davidson, “Wind vector retrieval using ERS-1 synthetic aperture radar imagery,” IEEE Trans. Geosci. Remote Sens., vol. 34, no. 6, pp. 1343–1352, Nov. 1996.
[5] M. Portabella, A. Stoffelen, and J. A. Johannessen, “Toward an optimal inversion method for synthetic aperture radar wind retrieval,” J. Geophys. Res., Oceans, vol. 107, no. C8, pp. 1-1–1-13, 2002.
[6] V. Kerbaol, B. Chapron, and P. W. Vachon, “Analysis of ERS-1/2 syn- thetic aperture radar wave mode imagettes,” J. Geophys. Res., Oceans, vol. 103, no. C4, pp. 7833–7846, 1998.
[7] V. Corcione, G. Grieco, M. Portabella, F. Nunziata and M. Migliaccio, “A novel azi- muth cut-off implementation to retrieve sea surface wind speed from SAR imagery,” IEEE Transactions on Geoscience and Remote Sensing, vol. 57, no. 6, pp. 3331-3340, 2018.

Most of the human world population lives along the coast. Therefore, their lives are heavily affected by the meteorological phenomena that characterize these areas. In that sense, coastal winds play a relevant role. Indeed, for example, the presence of sea breezes, katabatic flows and orographic winds in general, can strongly impact local micro climates and, for example, wind energy potential. Furthermore, they play a fundamental role in the generation of local ocean currents and in the dispersion of air pollutants.
As such, accurate and highly sampled coastal winds observations are of paramount importance to modern societies. In order to pursue this objective, scatterometer-derived wind vectors are potentially very useful, but first the excessively land contaminated radar footprints should be removed, while the low contaminated ones should be corrected by means of a land contribution ratio (LCR) based Normalized Radar Cross Section (NRCS) correction scheme. In addition, the NRCS noise should be carefully characterized in order to properly weight the backscatter NRCS measurements contributing to the wind field retrieval.
An assessment of the noise (Kp) affecting the NRCS measurements of the Seawinds scatterometer (onboard QuikSCAT) is carried out in this study. An empirical method is used to derive Kp (Kp'), which is then compared to the median of the Kp values (Kp'') provided in the Level 1B Full Resolution (L1B) file with orbit number 40651, dated 10 of April 2007, and the main differences are discussed. A sensitivity analysis aiming at assessing the presence of any dependencies with respect to (w.r.t.) different wind regimes, the kind of scattering surface, the scatterometer view and the polarization of the signal is carried out. In addition, the presence of any biases is assessed and discussed. Finally, a theoretical NRCS distribution model is proposed and validated against real measurements.
The main outcomes of this study demonstrate that H-Pol measurements are noisier than those V-Pol, for similar wind speed regimes. In addition, the noise decreases with increasing NRCS values, in line with the expectations. Furthermore, Kp' may largely differ from Kp'', especially for the peripheral measurements, with differences up to 20%. In particular, the Kp values provided for the outermost slices seem to be understimated, especially for what concerns the H-Pol measurements with indices 6 and 7. In addition, the Kp' values estimated over the sea surface are lower than those estimated at all scattering surface types. This trend is not seen for Kp'', for which the differences are almost absent. Furthermore, some inter slice biases up to 0.8 dB are present for H-Pol acquisitions while these only are up to 0.3 dB for V-Pol ones, in both cases increasing with the relative distance between the slices, in line with the general Geophysical Model Function (GMF) sensitivity as a function of incidence angle. These biases have a non-flat trend w.r.t. the acquisition azimuth angle for both polarizations. These small variations may be due to changes in the wind speed and direction distribution for each bin.
The theoretical NRCS distribution proves to be effective. It can be used both for simulation studies and for checking the accuracy of the NRCS noise.

Introduction
With the increased numbers of marine traffic and the man-made objects in the oceans, such as ships, oil platforms and many others, it becomes indispensable to detect these objects to benefit the maritime applications. The abundance of SAR data and its free-open availability encourage the researchers and the industry to exploit this unique remote sensing data to characterize different scatterers in oceans.
SAR wind retrieval processing depends mainly on the Bragg scattering mechanism[1], the scattering of radar pulses caused by centimeter-scale waves on top of long waves. It is possible to relate the Normalized Radar Cross Section (NRCS) values resulting from these small waves to wind speed measurements using geophysical model functions such as CMOD5[2]. Nevertheless, existence of any other type of scattering than Bragg scattering can negatively violate the accuracy of the retrieved wind speed from SAR. The wind speed accuracy matters for many applications, such as offshore wind energy applications. Therefore, quad polarized SAR data can be the key to improve the accuracy of SAR wind retrieval and create quality flags for the SAR wind maps based on different scattering mechanisms occurring in the imaged scene itself.
Different detection approaches can be used to characterize the anomalous pixels in SAR scenes. Constant False Alarm Rate (CFAR) is one of the prominent algorithms utilized to distinguish and detect ships in the ocean based on a threshold value. Nevertheless, this approach depends greatly on the background clutter distribution, subsequently; the CFAR algorithm may have severe problems at heterogeneous areas[3].

Datasets
Many satellite sensors are able to provide different polarization modes. Sentinel-1 (2014- present) collects C-band dual polarization over land worldwide as well as over priority coastal zones . PALSAR-1 (2006-2011) & 2 (2014-now), are side-looking phased array L-band SAR sensors and their Polarimetry mode (PLR) data are available. Last but not least, RADARSAT-1 (1995-2013) & 2 (2007-present) provide standard, wide and fine quad polarization mode. However, some challenges are facing these systems, among these, their technology is complex and the swath of the products is less compared with a single polarized system.

Theory
The diversity in full polarimetric datasets (POlSAR) allows us to complete characterization of the scattered objects rather than the dual and single polarized images, respectively. Each pixel in the full polarimetric scene can be represented in a scattering matrix (S), whereas its components are known as complex scattering measured as amplitudes from different combinations of V and H polarization, as follows:
S= [■(S_HH&S_HV@S_VH&S_VV )]
where S_HV is the backscattered coefficient from horizontal polarization transmission and vertical polarization reception. Other terms are similarly defined.
Scattering matrix gives information about the complete scattering process and can be directly employed to determine a deterministic or a single object scatter, but this is not the case we face near the offshore wind farms. Then, the S is random due to different scatterers that may exist in our study area. Speckle noise filtering is a crucial step in POlSAR processing to define accurately the covariance (C) and coherence matrix (T) while keeping the spatial resolution. Several general principles can be reviewed or implemented to select the best optimal principle, which fits our data. Not to mention all, one and multidimensional gaussian distribution, and the wishart distribution are used to study the distributed scatterers properties through estimation of the C and T matrix. In other words, the C and T matrix are obtained as a vectorization of the S matrix to obtain a new formulation to describe the information contained in the S matrix[4].

Methodology
This study is going to target different full and dual polarimetric datasets for offshore wind farms areas. The methodology of this study is illustrated as shown in the flowchart diagram. The methodology has the validation approach to validate the output of polarimetric decomposition outputs with conventional algorithms such CFAR, Likelihood ration test (LRT) for POlSAR ship detector, and faster Regional Convolution Neural Network(R-CNN) model. Furthermore, the non-Bragg scattering areas will be handled using a developed deep learning (DL) model to refill these areas with proper NRCS values to infer wind speed values.


Expected outcomes
Work is going on to approach the end product of the workflow: wind fields retrieved from SAR with added information about the wind speed quality. This research will definitely benefit many maritime applications and especially the offshore wind energy applications.

Acknowledgements
The PhD project belongs to the Train2Wind network. The Innovation Training Network Marie-Curie Actions: Train2Wind has received funding from the European Union Horizon 2020. Thanks to the European Space Agency for providing us with the full polarimetric datasets.

References
[1] G. R. Valenzuela, “Theories for the interaction of electromagnetic and oceanic waves - A review,” Boundary-Layer Meteorol., vol. 13, no. 1–4, pp. 61–85, 1978, doi: 10.1007/BF00913863.
[2] H. Hersbach, “Comparison of C-Band scatterometer CMOD5.N equivalent neutral winds with ECMWF,” J. Atmos. Ocean. Technol., vol. 27, no. 4, pp. 721–736, 2010, doi: 10.1175/2009JTECHO698.1.
[3] C. Liu, P. W. Vachon, R. A. English, and N. Sandirasegaram, “Ship detection using RADARSAT-2 Fine Quad Mode and simulated compact polarimetry data,” no. February, p. 74, 2010.
[4] Y. Yamaguchi, Polarimetric Synthetic Aperture Radar. 2020.

High resolution accurate coastal winds are of paramount importance for a variety of applications, both civil and scientific. For example, they are important for monitoring some coastal phenomena such as orographic winds, coastal currents and the dispersion of atmospheric pollutants, or for the deployment of off-shore wind farms. In addition, they are fundamental for improving the forcing of regional ocean models and, consequently, the forecasting of some extreme events such as the Acqua Alta often occurring in the Venice lagoon.
Scatterometer-derived winds represent the golden standard. However, their use in coastal areas is limited by the land contamination of the backscatter Normalized Radar Cross Section (NRCS) measurements. Nonetheless, the coastal sampling may be improved if the Spatial Response Function (SRF) orientation and the land contamination are properly considered in the wind retrieval processing chain.
This study focuses on improving the coastal processing of the Seawinds scatterometer onboard QuikSCAT as part of a EUMETSAT study in the framework of the Ocean Sea Ice Satellite Application Facilities (OSI-SAF).
In particular, the analytical model of the SRF is implemented with the aim of computing the so-called Land Contribution Ratio (LCR), which is, by definition, the portion of the footprint area covered by land. This index is then used for a double purpose: a) removing the excessively contaminated measurements; b) implementing a LCR-based NRCS correction scheme for the relatively low contaminated measurements. A second SRF estimate is obtained from a pre-computed Look-Up Table (LUT) of SRFs that are parameterized with respect to (w.r.t.) the orbit time, the latitude of the measurement centroid and the azimuth antenna angle.
Finally, the useful measurements (including those LCR-based corrected) are averaged in order to obtain integrated measurements by beam or view, which are then input in the wind field retrieval processor. Two different averaging procedures, i.e., a box car and a noise-weighted averaging, are implemented
A detailed comparison between the anaytical and the LUT-based SRF models is shown and the consistency of the derived LCR indices is verified against the coastline. A sensitivity analysis of the LCR-based NRCS correction scheme w.r.t. the LCR threshold is carried out. The effects of both averaging procedures on the retrieved winds are carefully analyzed. Finally, the retrieved winds are validated against some coastal buoys and their accuracy is assessed. Preliminary results will be presented and discussed at the conference.

The STP-H8 satellite mission, sponsored by the US Department of Defense (DoD), aims to demonstrate new low-cost microwave sensor technologies for weather applications. H8 will carry the Compact Ocean Wind Vector Radiometer (COWVR) and Temporal Experiment for Storms and Tropical Systems (TEMPEST) instruments to be launched and hosted on the International Space Station (ISS) in December 2021. This presentation will highlight the science and applications that can be enabled and enhanced by measurements from this mission. COWVR and TEMPEST together provide real-time, simultaneous measurements of ocean-surface vector winds (OVW), precipitation, precipitable water, water vapor profile and other atmospheric variables. Because of the ISS’ non-sun-synchronous orbit, these measurements will span across different times of the day for a given location. Similar to the capabilities provided by the ISS-RapidScat (2014-2016), COWVR’s OVW measurements can enhance the research of diurnal and semi-diurnal wind variability and facilitate the inter-calibration of OVW measurements from the sun-synchronous scatterometers. In particular, the OVW measurements from COWVR combined with those from the currently-operating satellite scatterometers make it feasible to estimate diurnal and semi-diurnal cycles of the OVW. Moreover, the simultaneous measurements of air-sea interface and atmospheric variables provided by COWVR and TEMPEST offer a unique opportunity to advance science and applications for weather, climate, and air-sea interaction. The real-time measurements from the mission are amenable for operational applications.

Preventing of high sea state generated in cyclonic conditions is a continous challenge for operational meteorological centers in order to ensure the best wave forecast in open ocean and coastal areas. The accuracy of wind forcing in such extreme conditions is important for capturing the best initial conditions for swell propagation over long distance. Recently the winds from SMAP and SMOS L-band radiometers have shown their ability to observe strong winds that exceed 40 m/s as in cyclonic conditions (Reul et al 2012). The objective of this study is to assess the impact of using radiometers winds on wave forecasting during cyclone seasons in north Atlantic and Pacific ocean and Indian ocean. The work consists of implementing an hybrid wind forcing composed of radiometers winds and model winds. A deep learning techinque is performed to ensure consistent cyclogenesis conditions. Simulations of the wave model MFWAM have been developed for cyclones and hurricanes cases, which caused severe damages after their passage (Bejisa, Lorenzo 2019, Hagibis 2019, etc.). The validation of model results has been performed with altimeters wave data. The results show a positive impact on significant wave height with good reduction of bias and scatter index. We also point out the good consistency between model and Sentinel-1 wave spectra near the cyclone trajectories.
In this study by using 1D ocean mixed layer model we investigated the impact of improved wind forcing on ocean circulation and key parameters such as temperature, currents and surface stress. Further results and comments will be presented in the final paper.

Over the ocean, an altimeter's measurements of normalised backscatter (sigma0) are interpreted as a measure of surface roughness, which is linked empirically with wind speed. However there are many other factors, both instrument specific and environmental, that affect the observed values. Observations at different radar wavelengths record the sea surface roughness at different spatial scales; I utilise the close relationship between backscatter strengths at the two most common altimeter frequencies (Ku-band and C-band) to highlight that other factors are present. Because of their differing sensitivities to scales of roughness, the sigma0 difference, sigma0_0Ku miinus sigma0_C, is not a simple offset, but has a peak for wind conditions around 6m/s. Instrumenting and processing options tender to alter the sigma0 values uniformly across a wide range of values, whereas environmental conditions tend to affect the shape of the relationship, as they are altering the interplay of different roughness scales or their radar-scattering properties.

As there is no universally recognised absolute calibration of altimeter sigma0 values, all instruments require a simple bias to bring them to a common scale. However there is also a dependence on the retracking algorithm used, with the Maximum Likelihood Estimator, MLE3, giving fairly robust estimates, whereas MLE4 and the standard SAMOSA retracker for SAR waveforms show a strong dependence on the inferred mispointing, and so need an adjustment correction for that. It has also been noted that the TOPEX and Jason altimeters in their non-sunsynchronous orbit have additional biases with a period of 58.77 days linked to the degree of solar exposure of the instrument in orbit.

There is an atmospheric effect that changes the sigma0_Ku values, and that is attenuation by liquid rain. This is only significant for a small percentage of observations, but the effect is more pronounced at the Ka-band of AltiKa and the future CRISTAL mission. The most marked environmental effect is due to wave height. At very low winds, a change in wave height of 1m can affect sigma0 by ~0.15 dB, but this causes minimal bias in wind speed estimates due to the low sensitivity to sigma0 of all wind speed algorithms in this regime. There is also an effect at high wind speeds which remains to be accurately quantified. Finally the sigma0_Ku-sigma0_C relationship appears to shift by about 0.15 dB on moving from tropical to polar waters. This confirms a previously reported temperature effect, although the size of the change is a little less than theoretically expected.

All these factors are reviewed within the scope of efforts to remove biases in wind speed algorithms that are either regional in nature or vary with satellite, so as to further efforts to develop a homogeneous altimetric wind speed product.


Figure shows the mean sigma0_Ku-sigma0_C relationships for 4 current altimetric satellites. On the left are the curves for each after initial bias adjustments, showing a divergence in behaviour at very high sigma0 (low winds); on the right are the variations of each curve with sea surface temperature.

Scatterometer data over the ocean are assimilated, at ECMWF and other NWP centres, in the form of ambiguous near surface ocean winds vector information. What scatterometer really measures is the surface radar backscatter that is essentially related to the directional roughness of the sea surface, which is fundamentally driven more by the surface stress caused by the relative motion between the atmospheric wind and the underlying ocean rather than the wind itself. Due to the lack of accurate in-situ surface stress measurements over the ocean to validate and calibrate scatterometer measurements, historically the scatterometer observations have been interpreted and calibrated to (equivalent neutral) wind rather than wind stress.
An ongoing EUMETSAT-funded project at ECMWF is investigating how to further increase the value of scatterometer observations in Numerical Weather Prediction (NWP), by assessing (and implementing) the assimilation of the surface stress, rather than wind components. In a coupled ocean-atmosphere data assimilation system, the ASCAT measurements assimilated as surface stress, can in principle, provide information on the atmospheric wind simultaneously constraining the ocean circulation.
An intermediate assimilation approach known as “stress equivalent winds” (following de Kloe et. al, 2017) is also being explored. This approach includes the sensitivity to air density variations.
A change in the type of assimilated observation variable required an adaptation of the observation operator and of the tangent linear and adjoint codes to enable the minimization in the 4D-Var analysis. New stress and stress equivalent wind observation operators have been developed (together with the tangent linear and adjoint) and tested and integrated into the Integrated Forecasting System (IFS). The error statistic assigned to the observations in the 4D-Var has been revised, a wind dependent error formulation has been characterized to be assigned to ASCAT surface stress observations. NWP observing system experiments assimilating ASCAT observations as either winds, stress or stress equivalent winds are being performed. Also the impact of ocean currents into the assimilation of ASCAT observations is under investigation.
The results of the study will be presented and discussed.

The University of Miami’s Center for Southeastern Tropical Advanced Remote Sensing acquired over 100 Synthetic Aperture Radar (SAR) images of the California Monterey Bay region for the ongoing Coastal Land-Air-Sea-Interaction Project. Approximately 30 of these images include signatures of nonlinear internal waves (NIW). Eight Air Sea Interaction Spar (ASIS) buoys deployed in the region of interest provide field measurements within the SAR image swath. Surface roughness is most commonly thought of as a result of the wind blowing over the ocean surface. SAR senses the short-scale ocean surface roughness by means of Bragg scattering. Although internal waves are subsurface waves, they are visible in SAR data because they modulate the surface currents resulting in increased roughness associated with the leading edge of the internal wave and decreased roughness associated with the trailing edge of the internal wave. Changes in surface roughness alter the drag coefficient which is a key parameter for detecting wind stress. It has been speculated that NIWs can drive wind velocity and stress variance relative to the mean atmospheric flow, suggesting a surface roughness—wind feedback mechanism exists.

Using the SAR images to confirm the presence of NIWs, we estimate the likely time of arrival at an ASIS buoy site if not already intersecting with a buoy at the time of the image acquisition. The ultrasonic anemometer mounted on ASIS provides the three components of velocity needed to derive the turbulent fluctuations of wind velocity, along with the product term u’w’. The Morlet wavelet transform is used to decompose the signal into both frequency and time domain to study the evolution of features. The length scales corresponding to a particular frequency band of enhanced energy patterns found in wavelet plots are compared to SAR measured NIW wavelengths. We take the covariance between u’ and w’ and integrate over frequency to see if this proposed NIW-induced change in wind stress occurs over the same particular frequencies. To assess the contribution of NIWs to the total air-sea flux, we take the cumulative cospectral sum of u’ and w’ (components of momentum flux). A neighboring ASIS buoy not in the path of an NIW is used to represent the background atmospheric flow. We will present early results and discuss the implications NIWs have on the momentum flux and if they should be considered when studying fine-scale ocean-atmosphere interactions.

Swells are long-crest waves induced by storms. They can travel thousands of kilometers and impact remote shorelines. They also interact with local wind generated waves and currents. It has been shown that the presence of swell lowers the quality of the geophysical parameters which can be retrieved from the delay/Doppler radar altimeter data. This, in turn, affects the estimation of small-scale ocean dynamics. In addition, the resolution offered by the delay/Doppler processing schemes, which is approximately 300 m spacing in the along-track direction, does not allow to resolve swells. This work presents a method which demonstrates that Synthetic Aperture Radar (SAR) altimeters show potential to retrieve swell-wave spectra from fully-focused SAR altimetry processed data for the first time, and proposes thus, that SAR altimetry can serve as a source for swell monitoring.

We present the first spectral analysis of fully-focused SAR altimetry data with the objective of studying backscatter modulations caused by swell. Swell waves distort the backscatter in altimetry radargrams by means of velocity bunching, range bunching, tilt and hydrodynamic modulations. These swell signatures are visible in the trailing edge of the waveform, where the effective cross-track resolution is a fraction of the swell wavelength. By locally normalizing the backscatter and projecting the waveforms on an along-/cross-track grid, satellite radar altimetry can be exploited to retrieve swell information. The fully-focused SAR spectra are verified using as reference buoy-derived swell-wave spectra of the National Oceanic and Atmospheric Administration's buoy network. Using cases with varying wave characteristics, i.e., wave height, wavelength and direction, we present the observed fully-focused SAR spectra, relate them to what is known from side-looking SAR imaging systems and adapt it to the near-nadir situation. Besides having a vast amount of additional data for swell-wave analysis, fully-focused SAR spectra can also help us to better understand the side-looking SAR spectra.

The study presents a method and application for estimating series of integrated sea state parameters from satellite-borne synthetic aperture radar (SAR), allow processing of data from different satellites and modes in near real time (NRT). The developed Sea State Processor (SSP) estimates total significant wave height SWH, dominant and secondary swell and windsea wave heights, first, and second moment wave periods, mean wave period and period of wind sea. The algorithm was tuned and applied to the Sentinel-1 (S-1) C-band Interferometric Wide Swath Mode (IW), S-1 Extra Wide (EW) and S-1 Wave Mode (WM) Level-1 (L1) products and also to the X-band TerraSAR-X (TS-X) StripMap (SM) and TS-X SpotLight (SL) modes. The scenes are processed in a spatial raster format and result in continuous sea state fields. However, for S-1 WV, the averaged values for each sea state parameter are provided for each 20 km×20 km imagette acquired every 100 km along the orbit.
The developed empirical algorithm consists of two parts: CWAVE_EX (extended CWAVE) based on widely known empirical approach and additional machine learning postprocessing. A series of new data preparation steps (i.e. filtering, smoothing, etc.) and new SAR features are introduced to improve accuracy of the original CWAVE. The algorithm was tuned and validated using two independent global wave models WAVEWATCH-3 (NOAA) and CMEMS (Copernicus) and National Data Buoy Center (NDBC) buoy measurements. The reached root mean squared errors (RMSE) for CWAVE_EX for the total SWH are 0.35 m for S-1 WV and TS-X SM (pixel spacing ca. 1–4 m) and 0.60 m for low-resolution modes S-1 IW (10 m pixel spacing) and EW (40 m pixel) in comparison to CMEMS. The accuracies of the four derived wave periods are in the range of 0.45–0.90 s for all considered satellites and modes. Similarly, the dominant and secondary swell and wind sea wave height RMSEs are in the range of 0.35–0.80 m compared to CMEMS wave spectrum partitions. The postprocessing step using machine learning, i.e., the support vector machine technique (SVM), improves the accuracy of the initial results for SWH. The resulting accuracy of SWH reaches an RMSE of 0.25 m by SVM postprocessing for S-1 WV validated using CMEMS. Comparisons to 61 NDBC buoys, collocated at distances shorter than 100 km to S-1 WV worldwide imagettes, result into an RMSE of 0.31 m. All results and the presented methods are novel in terms of achieved accuracy, combining the classical approach with machine learning techniques. An automatic NRT processing of multidimensional sea state fields from L1 data with automatic switching for different satellites and modes was also implemented. The algorithms provide a wide field for applications and implementations in prediction systems.
The SSP is designed in a modular architecture for S-1 IW, EW, WV and TS-X SM/SL modes. The DLR Ground Station “Neustrelitz” applies the SSP as part of a near real-time demonstrator service that involves a fully automated daily provision of surface wind and sea state parameters estimates from S-1 IW images of the North and Baltic Sea. Due to implemented parallelization, a fine raster for scene processing can be applied: for example, S-1 IW image with large coverage of around 200 km×250 km can be processed using a raster of 1 km (~50,000 analyzed subscenes) within few minutes.
The complete archive of S-1 WV L1 Single Look Complex (SLC) products from December 2014 until February 2021 was processed to create a sea state parameter database (121,923 S-1 WV overflights with around 3,000 IDs/months, each overflight consisting of 40-180 imagettes, total around 14 Mio S-1 WV imagettes). All processed S1 WV data including derived eight state parameters, quality flag and imagette information (geo-location, time, ID, orbit number, etc.) are stored as ascii and in netcdf format for convenient use. The derived state parameters are available to the public within the scope of ESA’s climate change initiative (CCI).
The validation carried out for the whole S-1 WV archive using CMEMS sea state hindcast for latitudes of -60° < LAT < 60° to avoid ice coverage with around 13,5 Mio collocations resulted in an RMSE of 0.245/0.273 m for wv1/wv2 imagettes, respectively. The SWH accuracy for different wave height domains for wv1/wv2 is as follows: 0.28/0.34 m (SWH < 1.5 m), 0.19/0.22 m (1.5 m < SWH < 3 m), 0.30/0.33 m (3 m < SWH < 6 m) and 0.51/0.55 m (SWH > 6 m). The monthly estimated total RMSE varies form 0.22 m to 0.31 m. These RMSE fluctuations around the mean value are caused by different amounts of acquired storms in individual months. As high waves have a higher RMSE, they increase the total RMSE when their relative percentage in a month is higher: in total, SWH distribution in the worldwide acquired SAR data is SWH < 3 m for ~75% of all cases, 3 m < SWH < 6 m for around 24% and only around 1% for SWH > 6 m and even less than 0.1% for SWH > 10 m. However, SWH > 6 m can reach around 2% for individual months with quadratic impact of the SWH values on RMSE.
The cross validations carried out using CMEMS, WW3 and mixed CMEMS/WW3 ground truth show: in terms of total SWH, in comparison to NDBC data, using only CMEMS ground truth resulted into an accuracy ~3 cm better than when the model function was tuned using WW3 data. This might be consequence of the better CMEMS spatial model resolution of 1/12 degree in comparison to WW3 (1/2 degree, spatially interpolated). The SWH comparison between CMEMS, WW3 and NDBC resulted into an RMSE=0.26 cm for CMEMS/NDBC and an RMSE=0.23 cm for CMEMS/WW3 at NDBC buoy locations. Generally, in terms of SWH, the ground truth noise can be assessed to an error of ~0.25 m. As can be seen, the resulting RMSE of 25 cm for S1 WV brings the results down to the noise level of the ground truth data.

Swells are waves from other ocean areas or generated locally but do not absorb energy from the wind anymore. Swells have longer wavelength than wind waves and can propagate over a very long distance in the ocean. In this study, the in-situ data from NDBC buoys located openly to the southwest are used to determine the potential destination of swells propagating from the westerlies for southern hemisphere. Meanwhile, the CFOSAT SWIM data and ST6 reanalysis data are used to trace back the trajectories and the sources of these swells. Accordingly, we find 25 routes of swells originated from 4 series of ocean storms. To verify the accuracy of these paths, we check the variation of wave parameters in 48 hours before and after SWIM observation. It shows clearly that swells from the southwest have passed through and continue to travel northeastward. The one-dimensional wave spectra of SWIM data and NDBC buoy data are compared which indicates the attenuation of energy. It is shown that magnitude of decaying rate of swell energy increases with the spectral width of initial swell field. In addition, the general rate of increase for peak wavelength is an order of 0.01m/km, which is apparently the spectral width dependent. These are mainly due to the higher degree of dispersion and angular spreading for broader spectra. To quantity the energy that decays due to spherical spreading, the point source hypothesis is used between SWIM observation points and NDBC buoys. Besides, the ST6 reanalysis dataset without considering swell dissipation and negative wind input is compared to the real observation data to help obtain the spherical spreading values from sources to SWIM and from sources to buoys. Linear and nonlinear dissipation rates are calculated according to the air-sea interaction theory and wave-turbulence interaction theory. The result shows that the intensity of dissipation is stronger near the sources (the linear dissipation rate is about 10^(-7) m^(-1)) and decreases in the subsequent propagation.

One of the challenge of future earth system for climate prediction, is to better understand the exchange of momentum, heat and gas fluxes at air-sea interface. In this context, the waves play a key role for the estimate of accurate forcing terms to the upper ocean layers and feedback to the atmosphere. Currently the global wave system of the Copernicus Marine Service (CMEMS) is jointly assimilating Significant Wave Height (SWH) from altimeters and directional wave observations from CFOSAT and Sentinel-1. This leads to significant improvement of integrated wave parameters of sea state, particularly in ocean regions affected by strong uncertainties related to the wind forcing, for instance the Southern Ocean. Among the promising recent developments, we can highlight the synergy between waves and wind observations as provided by CFOSAT mission, which has shown the capacity of retrieving wide swath SWH with good accuracy (Wang et al. 2021). This work gives an overview of using both wide swath SWH and directional wave spectra from satellite missions (CFOSAT, Sentinel-1, HY-2B, HY-2C) in operational wave model. We show that the performance of such assimilation system induces a significant reduction of normalised scatter index of SWH, which is in average smaller than 8%.
We investigated the impact of using both directional wave spectra and wide swath SWH in critical ocean regions suh as the southern ocean and the tropics, with particular attention to consequences on ocean circulation. We also draw up the improvement induced by the assimilation of directional wave spectra on the wind-wave growth and the estimate of wave group speed under unlimited fetch conditions. In this work we examined the complementary use of wave spectra from wave scatterometter SWIM and SAR for better capturing swell propagation. In other respects the persistency of the assimilation of wide swath SWH and directional wave spectra is extended to 4 days in the forecast period, which ensures good reliability for wave submersion warning and marine security bulletins.
Further results concerning the impact of wave directionality on upper ocean mixing has been investigated in the tropics and in the area of Antarctic Circumpolar Current (ACC) ocea area. figure illustrates the zonal mean of eastward component of the current between 146°-149°E of longitudes in Southern ocean from NEMO model simulations and observations from drifters (AOML products). This clearly reveals the improvement of the surface currents when ocean model is coupled with improved wave forcing which uses directional wave spectra and swath SWH.
More discussions and conclusions will be summarized in the final presentation.

Inspired on the work of Cavaleri et al. [2012], where the concept of the so-called swell-wind (i.e., a "low-level wave-driven wind jet" as described in 2010 by Hanley et al.) is discussed, this work-in-progress aims to investigate the spatial variability of wind-wave coupling in a semi-enclosed basin through the analysis of SAR imagery. A global climatology and seasonal variability of wind-wave coupling was computed by the latter authors through the inverse wave age parameter, derived from numerical results of the ERA-40 dataset. They identified areas where, and times when, wind-driven wave conditions (U_10 cosθ / c_p > 0.83) and wave-driven wind regimes (U_10 cosθ / c_p < 0.15) occur, the latter coinciding with the swell pools found by Chen et al. [2002]. They also found that in enclosed seas, where wave-growth (and hence, c_p) is limited by fetch, the wind-driven wave regime is most common, as proposed by Drennan et al. [2003].

In this work, maps of inverse wave age have been computed from wind and wave information derived from Sentinel-1 and TerraSAR-X images acquired over the Gulf of Mexico and validated with buoy observations. Mean wave conditions at the Gulf of Mexico are mild (Hs < 2 m) and mostly driven by the Trade Winds, so that its propagation direction has a strong zonal component and is somewhat normal to these platform's flight direction. This allows for a reliable SAR detection. Under such conditions, inverse wave age lies in the "mixed" regime (0.15 < U_10 cosθ / c_p < 0.83) and its spatial variability appears to be induced mostly by the wind's. However, when the sea surface is forced by the high-wind conditions associated with atmospheric cold surges (November through May) and tropical cyclones (June through November), waves can be as high as 7 m and reach peak period values above 13.5 s, as recorded by NDBC stations 42002 and 42055 in 2020. Spatial variability of the inverse wave age parameter seems then to be dependent mostly on the phase speed and propagation direction of the waves, specially in cases where shorter- and longer-wavelength systems coexist. During these extreme conditions, both wind-driven seas and wave-driven winds have been estimated, indicating areas where the ocean yields momentum into the atmosphere. This study supports the hypothesis that momentum flux can be highly variable (i.e., spatially inhomogeneous) not only near the coast but in the open ocean, as proposed by Laxague et al. [2018].

Global long-term wave climate models are essential to estimate changing climate impacts on future projected sea states, which are crucial for offshore safety and coastal adaptation strategies. In such projections, wave climate models are forced with Global Circulation Model (GCM) wind speed and sea-ice concentration to simulate the wind-wave evolution over extensive time scales. However, GCMs are affected by external forcing and internal variability uncertainties. As such, a model democracy approach, where each model equally contributes to the analysis of the future projected wind-wave climate may result in a high spread in future projection estimates that, if averaged from global statistics, could mask stronger signals in the ensemble best performing models (Knutti et al., 2017). The common practice to overcome such constraints is to use bias-corrected or weighted wave climate model ensembles to estimate the average past and future climate (Morim et al., 2019, Meucci et al., 2020). This work describes a novel observation-based weighting approach based on an in-detail assessment of CMIP6 and CMIP5 derived wave climate model performance using a 33-year calibrated satellite dataset (Ribal and Young, 2019). We compare the wave climate model statistics with collocated satellite measurements at the global level and selected climatic regions (Iturbide et al., 2020). We evaluate the mean climatology, trends and extreme wave estimates of each model. The models are then classified using the Knutti et al. (2017) weighting formula that considers model performance and interdependence. The result is a wave climate ensemble weighted by global observational statistics which should serve as an optimally balanced dataset for future ensemble statistical studies and Extreme Value Analyses ensemble approaches (Meucci et al., 2020).

References:

Iturbide, M., Gutiérrez, J. M., Alves, L. M., Bedia, J., Cerezo-Mota, R., Cimadevilla, E., ... & Vera, C. S. (2020). An update of IPCC climate reference regions for subcontinental analysis of climate model data: definition and aggregated datasets. Earth System Science Data, 12(4), 2959-2970.

Knutti, R., Sedláček, J., Sanderson, B. M., Lorenz, R., Fischer, E. M., & Eyring, V. (2017). A climate model projection weighting scheme accounting for performance and interdependence. Geophysical Research Letters, 44(4), 1909-1918.

Meucci, A., Young, I. R., Hemer, M., Kirezci, E., & Ranasinghe, R. (2020). Projected 21st century changes in extreme wind-wave events. Science advances, 6(24), eaaz7295.

Morim, J., Hemer, M., Wang, X. L., Cartwright, N., Trenham, C., Semedo, A., ... & Andutta, F. (2019). Robustness and uncertainties in global multivariate wind-wave climate projections. Nature Climate Change, 9(9), 711-718.

Ribal, A., & Young, I. R. (2019). 33 years of globally calibrated wave height and wind speed data based on altimeter observations. Scientific data, 6(1), 1-15.

Long-period swells generated in the North and South Pacific frequently hit the shores of low-lying Pacific islands and atolls. The accuracy of wave forecasting models is key to efficiently anticipate and reduce damage during swell-induced flooding episodes. However, in such remote areas, in situ spectral wave observations are sparse and models are poorly constrained. Therefore, Earth Observation satellites monitoring sea state characteristics, represent a great opportunity to improve the forecasting of flooding episodes, and the analysis of wave climate variability.

Here, we present a satellite-driven swell forecast system that can be applied worldwide to predict the arrival of swells. The methodology relies on the dispersive behavior of ocean waves, assuming that the energy travels along great circle paths with a celerity that only depends on its frequency. Satellite data for this analysis are directional wave spectra derived from SWIM acquisitions onboard the CFOSAT-mission (Hauser et al. 2021).

The proposed workflow includes: a) filtering the global-coverage data considering a temporal and geographical criterion (the spatial scale delimits the effective energy source that can reach the target location); b) comparison of wave parameters from CFOSAT-SWIM and partitions from a global wave numerical model for the removal of the directional ambiguity; c) definition of the spectral energy sector that points towards the study site; d) analysis of air-sea fluxes dissipation (Ardhuin et al. 2009); and, e) analytical propagation of the energy bins to forecast the targeted spectral energy over time.

Two examples of application are presented for the Samoa islands, and for the Cantabria coast (Spain). The time evolution of swell systems approaching the sites is evaluated against spectral energy from available in situ wave measurements and numerical model outputs. The results exhibit a good reproduction of the wave fields, proving the flexibility and robustness of the methodology.

The proposed method may be used to track swells across the ocean, forecast the arrival of swells or locate remote storms. For the Small Island Developing States, the output of the methodology can be undoubtedly of great help for stakeholders and decision makers to produce risk metrics
and implement strategies that minimize the vulnerability of these communities to coastal flooding at a very low computational effort.

ACKNOWLEDGEMENTS
This work was supported by the French National Research Agency through the ISblue program, (ANR-17-EURE-0015) and by CNES through the CFOSAT-COAST project.

REFERENCES
Ardhuin, F., Chapron, B., & Collard, F. (2009). Observation of swell dissipation across oceans. Geophys. Res. Lett., 36 , L06607. doi: 10.1029/2008GL037030

Hauser, D. et al., (2021). New Observations From the SWIM Radar On-Board CFOSAT: Instrument Validation and Ocean Wave Measurement Assessment. IEEE Transactions on Geoscience and Remote Sensing 59, 5–26. https://doi.org/10.1109/TGRS.2020.2994372

Ocean surface waves are modified by surface currents. This has strong implications for remote sensing of wind and currents by classical or Doppler scatterometry, especially at high horizontal resolution.
We discuss here different mechanisms of wave modification around a current front. In particular, we compute propagation and dissipation effects using a numerical wave model of wave action conservation. We show that short wind waves, long wind waves and long non-dissipative swell all respond differently to the different current gradient components. The horizontal scales and the degree at which those 3 responses can be coupled to each other is a key to understand this complex response of the wave field to currents.

Detailed knowledge of the shape of the seafloor is crucial to humankind. In an era of ongoing
environmental degradation worldwide, bathymetry data (and the knowledge derived from it) play
a pivotal role in using and managing the world’s oceans. Bathymetric surveys are used for many
research fields including flood inundation, the contour of streams and reservoirs, water-quality studies,
planning the coastal reservoir, and many other applications.
However, the vast majority of our oceans is still virtually unmapped, unobserved, and unexplored.
Only a small fraction of the seafloor has been systematically mapped by direct measurement.
For understanding changes of the underwater geomorphology, regional bathymetry information is
paramount. This sparsity can be overcome by space-borne satellite techniques to derive bathymetry.
With the development of new missions in open-access, space-borne sensors now represent an attractive
solution for a broad public to capture local-scale coastal impacts at large scales.
Only from intermediate water until shore, the linear dispersion relation (1) can be used to estimate
a local depth.
c² =g/h tanh( h/λ ) ⟺ h = λ atanh( c²/gλ ) (1)
in which c is wave celerity, g represents the gravitational acceleration, and λ is the wavelength.
Studies, for example [2], show that wave pattern can be extracted using a Radon transform then they
obtained physical wave characteristics (λ, c) using a 1D-DFT for the most energetic incident wave
direction in Radon space (sinogram).
In this work, we seek a thorough research in signal processing that is contained in Sentinel-2 (ESA/
Copernicus spaceborne optical sensor) images and optimization of this signal. This work is carried out
in the perspective of the production of differential bathymetry, with interest for detection / evaluation
of changes on underwater geomorphology. Identification of such changes has potential applications
in risk analysis related to seismotectonics, submersion, submarine gravitational movements and morphodynamics,
littoral dynamic-related seasonal or extreme event, among others.
Here, regional bathymetries are derived at the test-site of Arcachon, France.
Our approach is based on the calculation of the gradient around each point of the image. This approach
will be a source of improvement of the method [2, 4] and will give us a better estimation of
wave propagation direction and the possibility of dealing with two wave regimes which overlap.
When analyzing directional data, it is often appropriate to pay attention only to the direction of
each datum, disregarding its norm. The von Mises–Fisher (vMF) distribution is the most important
probability distribution for such data[3].
With this novel technique, we extract the wave direction by estimating the parameters of Von Mises-
Fisher distribution from local gradients around each point[1]. Therefore, Sentinel-2 imagery derived
wave characteristics are extracted using a unidirectional radon transform. A discrete fast-Fourier
(DFT) procedure in Radon space (sinogram) is then applied to derive wave spectra. Sentinel-2
time-lag between detector bands is employed to compute the spectral wave-phase shifts. Finally, we
estimate depth using the gravity wave linear dispersion equation (1).
In conclusion, the development of the theoretical model based on von Mises–Fisher(vMF) distribution
is an alternative way to carry out the processing in order to produce coastal bathymetry suggesting
potential improvements respect to previous approaches. Ultimate goals are to be able to make accurate
developments of our approach with the intention of improving the method to detect mixture of
von Mises-Fisher distributions.

Keywords— bathymetry, signal processing, spaceborne imagery

References
[1] Akihiro Tanabe, Kenji Fukumizu, Shigeyuki Oba, Takashi Takenouchi, and Shin Ishii. 2007. ”Parameter
estimation for von Mises–Fisher distributions” Computational Statistics 22(1), 145-157.
[2] Bergsma, Erwin W.J., Rafael Almar, and Philippe Maisongrande. 2019. ”Radon-Augmented Sentinel-2 Satellite
Imagery to Derive Wave-Patterns and Regional Bathymetry” Remote Sensing 11, no. 16: 1918.
[3] Lu Chen, Vijay P. Singh, Shenglian Guo, Bin Fang, and Pan Liu. 2012. ”A new method for identification of
flood seasons using directional statistics” Hydrological Sciences Journal 58(1), 1–13.
[4] Marcello de Michele, Daniel Raucoules, Deborah Idier, Farid Smai, and Michael Foumelis. 2021. ”Shallow
Bathymetry from Multiple Sentinel 2 Images via the Joint Estimation of Wave Celerity and Wavelength”
Remote Sensing 13, no. 12: 2149.

We retrieve significant ocean surface wave heights in the Arctic and Southern Oceans from CryoSat-2 data. We use a semi-analytical model for an idealised synthetic aperture satellite radar or pulse-limited radar altimeter echo power. We develop a processing methodology that specifically considers both the Synthetic Aperture and Pulse Limited modes of the radar that change close to the sea ice edge within the Arctic Ocean. All CryoSat-2 echoes to date were matched by our idealised echo revealing wave heights over the period 2011-2019 (Updated to 2021). Our retrieved data was contrasted to existing processing of CryoSat-2 data and wave model data, showing the improved fidelity and accuracy of the semi-analytical echo power model and the newly developed processing methods. We contrasted our data to in-situ wave buoy measurements, showing improved data retrievals in seasonal sea ice-covered seas. We have shown the importance of directly considering the correct satellite mode of operation in the Arctic Ocean where SAR is the dominant operating mode. Our new data is of specific use for wave model validation close to the sea ice edge and is available at http://www.cpom.ucl.ac.uk/ocean_wave_height/.

In the frame of the second phase of the Copernicus Marine Environment Monitoring Service (CMEMS), starting in 2022, the WAVE Thematic Assembly Centre (TAC), a partnership between CLS and CNES, is responsible for the provision of a near-real-time wave service that started in July 2017. Near-real-time wave products derived from altimetry and SAR measurements are processed and distributed onto the CMEMS catalogue.

This presentation will describe the existing products – along-track Level 3 and gridded Level 4 – and their applications such as near-real-time assimilation in wave forecasting systems, validation of wave hindcasts, etc.

Early 2022, Sentinel-6 will integrate the existing altimetry constellation measuring significant wave heights (SWH) and collocated wind speed. Sentinel-6 will become the reference mission of the CMEMS L3 SWH product, succeeding Jason-3 once it changes orbit. The Sentinel-6 Level-2P and Level-3 processing is under EUMETSAT and CNES responsibility and is operated by CLS. We will describe the processing steps from Level-2 to Level-3 product carried out to produce a homogeneous dataset with regard to other WAVE-TAC altimetry datasets.

The daily gridded Level-4 SWH product will also benefit from the integration of Sentinel-6. The increased spatial and temporal density of measurements will allow a better mapping of the wave heights.

In the frame of this presentation, we will produce a thorough comparison of CMEMS Level-3 & Level-4 SWH products versus in-situ measurements provided by the In-Situ TAC. In particular, we will highlight the changes induced by the integration of the new Sentinel-6 mission in the WAVE-TAC product.

Abstract:

The spectral characteristics of SLC-IW TOPS are significantly different from Strip-map (SM). Due to the burst mode and series of sub-swaths, the target area is scanned for a short period of time, consequently, swath width comes at the expense of azimuth resolution. Significant focusing is required to remove quadratic phase drift and achieve an SLC base-band. The de-ramping effectively eliminates the quadratic drift and restores the baseband data. The ocean circulation parameters are extracted from the echo signal based on data-driven Doppler centroid (DC). The ocean circulation parameters include surface velocity, wave height, and direction swell, while compared with the synergy of benchmark data.

Background:

Due to burst-mode, SLC-IW TOPS differs from SM in terms of schematics, and the system observes in the form of sub-swaths periodically. As a result, the target region is scanned only for a fraction of the burst duration, and thereby the illumination is reduced, and the wide swath comes at the cost of azimuth resolution [1]. Sentinel-1 IW TOPS data preserves quadratic phase term in the azimuth direction which
leads to phase ramps, this needs to be eliminated from the SLC data for the subsequent applications.
In literature, the ocean circulation parameters for SLC-IW data are estimated based on the information provided in the OCN level-2 product, or geophysical interpretation is calculated from satellite orbit parameters [2]. The orbit parameters velocity V and incident angle θ in practical is usually not accurate enough to obtain the DC which fulfills the need for SAR imaging. Therefore, this work estimates the DC and all associated parameters from echo data [3], and all the ocean circulation parameters are data-driven.

Methodology:

To remove the quadratic drift, it is essential to move the spectral component of SLC-IW to the baseband by deramping. The phase term for deramping is defined as:
ϕ(η, τ ) = exp{(−j.π.k_t(τ )) × (η − η_ref (τ ))^2} (1)
whereas, reference time η_ref (τ ), and Doppler centroid rate k_t(τ ) are functions of range samples, while η is zero-Doppler azimuth time. The phase term needs to be multiplied in time
domain with SLC signal S_slc.
S_d(η, τ ) = S_slc × ϕ(η, τ ) (2)
Alternatively, deramping can be done in the SNAP tool using the Sentinel-1 TOPS operator. The flow of the process is given in the figure.

On that account, the Doppler centroid is the essence of this topic. In the literature, the DC has been predicted by conventional OCN product information using DC polynomial information provided in the metadata. We use correlation doppler estimation (CDE) which takes an advantage of azimuth shift and the PRF [4]. This DC history is utilized to retrieve radial surface velocity (RSV) with incident angle and radar frequency information. And with the empirical relationship of RSV, we estimated significant wave height (SWH) [5]. The SWH is the average wave height (from trough to crest) of the highest third of the wave height during the sample period. The comparisons are made with the synergy of benchmark data (measured by OCN product) for the same location, date, and time while using the SLC-IW TOPS product [6].

Results and discussion:

The quadratic drift is removed when phase term ϕ(η, τ) is multiplied with the original SLC image the data moves to the baseband domain. Deramping is done so far to eliminate the quadratic effect of phase term by chirp signal. , we extract ocean circulation parameters for the post-processing. For this, we measure the RSV based on DC information estimated by the CDE method, which perfectly matches with benchmark data. The RSV is in a good match and within the limit of error bounds, while in the core of the stream it reaches up to 2.5 m/s.
The RSV is an associated term for retrieving significant wave heights (Hs), which varies by a few meters. we use dual-polarization VH, which provides a better estimate of Hs than single polarization.

Conclusion:

The designed chirp function de-ramps the data and the result is theoretically correct, whereas the data moves to the baseband. The ocean
circulation parameters are measured and numerical values are compared with benchmark data and found perfectly matched. The numerical merit of comparisons is in good spatial correlation with minimum root mean square error (RMSE) and negligible mean absolute error (MAE).

References:

[1]. De Zan, Francesco, and A. Monti Guarnieri. "TOPSAR: Terrain observation by progressive scans."IEEE Transactions on Geoscience and Remote Sensing, 44.9 (2006): 2352-2360.
[2]. Hansen, Morten Wergeland, et al. "Retrieval of sea surface range velocities from Envisat ASAR Doppler centroid measurements."IEEE Transactions on Geoscience and Remote Sensing, 49.10 (2011): 3582-3592.
[3]. Zou, Xiufang, and Qunying Zhang. "Estimation of Doppler centroid frequency in spaceborne ScanSAR."Journal of Electronics (China),25.6
(2008): 822-826.
[4]. M. Amjad Iqbal, Andrei Anghel, and Mihai Datcu, ”Doppler Centroid Estimation for Ocean Surface Current Retrieval from Sentinel-1 SAR
Data”, IEEE Eu-RAD conference European Microwave week, 2022.
[5]. Pramudya, Fabian Surya, et al. "Enhanced Estimation of Significant Wave Height with DualPolarization Sentinel-1 SAR Imagery." Remote Sensing, 13.1 (2021): 124.
[6]. AElyouncha, Anis, Leif EB Eriksson, and Harald Johnsen. "Comparison of the Sea Surface Velocity Derived from Sentinel-1 and Tandem-X." 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS. IEEE, 2021.

In the Sea State community, the literature usually assumes that altimetry waves at 1Hz is dominated by noise (Ardhuin et al. 2019) and most studies tackle this issue by filtering these data up to 50km at least (Quilfen et al. 2018, Dodet et al. 2020). It is also known that the fading noise has a real impact on correlated errors at these scales (Quartly et al. 2019), with an impact on SLA estimates that can empirically be reduced by methods described in Zaron et al. or Tran et al. 2021.

Is this presentation, we propose to process the 20Hz resolution altimetric data and to look deeper into this HF content. After a 5Hz compression, we analyze the frequential content and their geographical signatures. We also take a particular care to data selection, an essential step for validation purpose (as illustrated in Quartly et al. 2019).

The analysis is multimission oriented. LRM missions are processed with the innovative adaptive retracking (Tourain et al. 2021). It also deals with Doppler SAR altimetry observations, also impacted by the sea sate structures (Moreau et al. 2021). The study focuses on Jason3 thanks to SALP CNES project, ENVISAT, in the frame of the innovative Fdr4ALT project and CFOSAT (Hauser et al. 2020).

To better characterize the high frequency signal, we leverage the spectral information (direction and wavelength and partitions) provided by the CFOSAT mission, as well as Sentinel 1 radar imaging. A discussion is carried to determine how these altimetric products could become the next generation of CMEMS WaveTAC products. It also explores the contamination effects on SLA estimates, mainly in the bump frequency, extensively described in Dibarboure et al 2014.

The processing of these new 5Hz L2P products is presented. Their quality over coastal areas is illustrated and demonstrated. Their added value offshore is highlighted and offered to discussion with other international teams (for assimilation and/or validation, climate or coastal communities…) via demonstration products available on Aviso web site. Give it a try!

Sea state is a key component of the coupling between the ocean and the atmosphere, the coasts and the sea ice. Understanding how sea state responds to climate variability and how it affects the different compartments of the Earth System, is becoming more and more pressing in the context of increased greenhouse gas emission, accelerated sea level rise, sea ice melting, and growing coastline urbanization.

A new multi-mission altimeter product that integrates improved altimeter retracking and inter-calibration methods is being developed within the ESA Sea State Climate Change Initiative project. This effort will provide 30 years of uninterrupted records of global significant wave height. The 30 years represents the minimum duration required for computing climatological standards
following the World Meteorological Organization recommendation (WMO, 2015). In recent years, several
authors (e.g. Young et al., 2011; Young and Ribal, 2019; Timmermans et al., 2020) have computed the
trends in both the mean and extreme Hs using calibrated data from multi mission altimeter records. Whether these trends are the signature of anthropogenic climate change or the signature of natural variability is not known. Indeed, the atmosphere exhibits variability on a time scale comparable to the length of the satellite era and is therefore likely to hide the anthropogenic signal. Using the ECMWF ERA-5 reanalysis and focusing on North Atlantic region, we indeed show that the trends in winter-mean Hs computed over the satellite altimetry era are mostly associated with the atmospheric variability on the altimetry era time scale. Because the winter Hs variability in the North Atlantic is tightly linked with the overlying sea sevel pressure (SLP) winter variability, we extract the SLP modes of variability responsible for the altimetry era Hs winter trends in three regions (the Norwegian sea, the Mediterranean sea and the south of Newfoundland sea) where the Hs trends are significant.

In order to investigate the contribution of natural climate variability in these Hs trends, we analyzed SLP outputs obtained from the Community Earth System Model version 2 Large Ensemble (Lens2). Our analysis reveals that the magnitude of the SLP slope linked with internal variability becomes comparable to the magnitude of the slope linked with anthropogenic climate change for ~ 65 years of data i.e. around 2060, considering that 1992 (ERS-1 launch) is the beginning of the continuous altimetry era. This suggests that Hs modification associated with anthropogenic climate change of the atmospheric circulation will not be detectable in satellite altimetry trends before several decades.

IMOS (Integrated Marine Observing System) OceanCurrent (oceancurrent.imos.org.au) is a marine data visualisation digital resource that helps communicate and explain up to date ocean information around Australia. The information offers benefit to a broad range of users including swimmers, surfers, recreational fishers, sailors, and researchers using data collected from satellites, instruments deployed in the ocean, and accessible model outputs. The platform includes near-real-time data for sea surface temperature, ocean colour, and sea level anomaly from various satellite missions and in-situ instruments such as Argo, current meters, gliders, and CTDs etc. Until now, ocean surface wave information, both from in-situ wave rider buoys and satellite missions, has not been captured in OceanCurrent.

Australia has a growing network of moored coastal wave rider buoys. Network gaps are being identified (Greenslade et al., 2018, 2021) and filled, and new low-cost wave buoys are also being tested and deployed alongside traditional systems, further increasing the in-situ surface wave data captured. The publicly available national wave data network consists of approximately 35+ platforms operated by several different State and Commonwealth agencies and industry-contributed data (Greenslade et al., 2021). As it can be challenging and time-consuming to gather wave observations from various sources for large scale national or regional studies, IMOS AODN (Australian Ocean Data Network) Portal has strived to build an archive (and a near real time feed) of available national wave buoy observations. The AODN service is also being expanded by adding more platforms and by improving the meta-data of buoy record. Historical and near-real-time national wave data from a substantial set of wave buoys can now be easily accessed.

International satellite remote sensing radar altimeter and synthetic aperture radar (SAR) missions are also providing open data of surface wave observations globally. Also, CFOSat SWIM instrument has been providing global surface wave spectra measurements since launch in 2018 (Hauser et al., 2021). Using these valuable resources, Australia has developed, and continues to maintain, long-term multi-mission databases of calibrated wave height observations from radar altimeters (Ribal et al., 2019) and long-wave spectra from selected SAR missions (Khan et al., 2021). Some of these databases are also providing near real time feeds that can be exploited to gather up to date wave information.

An experimental national ocean surface waves product is under development for IMOS OceanCurrent Portal by integrating surface waves information from coastal buoys and satellite missions. As both radar altimeter and SAR satellites are polar orbiting with relatively narrow swaths (~10-20 km) over open ocean, during any short time window at best there are only a few along-track satellite measurements available. To convey a full representation of the wave field, background wave information from Bureau of Meteorology’s (BoM) AUSWAVE initialisation time step (t0) is shown. Surface wave maps are created at 2-hourly time steps with t0 as the central time showing AUSWAVE significant wave height and peak wave direction. Coastal buoy observations including significant wave height, mean wave direction, mean wave period, and directional spread within t0 +/- 3 hours, radar altimeter significant wave height within t0 +/- 1 hour, and peak wave direction and mean period extracted from SAR spectra within t0 +/- 30 mins are displayed, when available. Monthly videos from 2-hourly surface wave maps are also created to have a synoptic record of wave field propagation from ocean to the coast. The surface wave map archive currently spans 2021 and is planned to contain up to date (up to a few hours delay) surface wave information.

Once available on OceanCurrent, the hope is that this product will enable the wider community of recreational marine users and researchers to extract relevant surface wave information as needed and provide direct societal benefits by providing a national view of easily available and integrated surface wave information.

A sample image of the surface waves product is attached with the abstract to help reviewers, but it will likely be unavailable for the online abstract version (if accepted) as advised by the symposium organisers.

References

Greenslade, D. J. M., Zanca, A., Zieger, S., and Hemer, M. (2018): Optimising the Australian wave observation network. J. South. Hemisphere Earth Syst. Sci., 68, 184–200, https://doi.org/10.22499/3.6801.010.

Greenslade, D. J. M., Hemer, M. A., Young, I. R. & Steinberg, C. R. (2021). Structured design of Australia’s in situ wave observing network, Journal of Operational Oceanography, doi: 10.1080/1755876X.2021.1928394

Ribal, A., Young, I.R. (2019). 33 years of globally calibrated wave height and wind speed data based on altimeter observations. Sci Data 6, 77. https://doi.org/10.1038/s41597-019-0083-9

Khan, S. S., Echevarria, E. R., & Hemer, M. A. (2021). Ocean swell comparisons between Sentinel-1 and WAVEWATCH III around Australia. J. Geophys. Res: Oceans, 126, e2020JC016265. https://doi.org/10.1029/2020JC016265

D. Hauser et al., (2021). New Observations from the SWIM Radar On-Board CFOSAT: Instrument Validation and Ocean Wave Measurement Assessment," in IEEE Transactions on Geoscience and Remote Sensing, vol. 59, no. 1, pp. 5-26, Jan. 2021, doi: 10.1109/TGRS.2020.2994372.

The wave spectrum is a representation of the state of the ocean surface from which many parameters can be deduced: significant wave height, peak parameters of dominant waves, directional parameters, etc... For more than 30 years, Synthetic Aperture Radars allow their routine montoring far from the coast in all surface conditions (through clouds and despite night), where buoys cannot be deployed. These measurements benefit from scientific efforts that now make it a reliable measurement technique. Sentinel-1 constellation is one of them and operates since 2016. However known limitations, including wave blurring caused by the azimuth cut-off, limit the performance of wave spectra measurement to long swells.
SWIM is a new rotating Radar onboard the Chinese-French CFOSAT satellite dedicated to directional wave spectra measurement. Not suffering from the cut-off limitations, new horizons and perspectives for synergies are opened in terms of spectral limit and directionality.

Sentinel-1 wave spectra measurements are limited to Wave Mode acquisitions, only available over deep ocean basins and away from the North-East Atlantic Oean. SWIM measurements offer hundreds of co-located measurements over these regions, but also extends the coverage to closed seas worldwide and European waters. Other complementarities exist in wave measured wavelengths : Sentinel-1 extends to long swell (up to 800 m wavelength) while SWIM shows greater ability to measure wind sea components (close or below 50 m).

Those complementarities are assessed at different levels and with different comparisons methodologies. First, partition integral parameters are compared between SWIM Level-2P products or S1 Level-2 products and numerical wave model outputs alone. Second, dynamical co-locations are performed between S1 and SWIM, using cross-overs given by Level-3 spectral produtcs. These measurements, also referred to as Fireworks, allow to dramatically increase  the number of co-located points and better inter-compare.

These new performances have applications in data assimilation and prospects for new products such as Stokes drift, a first-estimated spaceborne measurement.

Harmony consists of two satellites that will fly in a constellation with one of the Sentinel-1 satellites. The two Harmony satellites carry a passive instrument that receives signals which are transmitted by Sentinel-1 and reflected from the surface. The full system therefore benefits from two additional lines-of-sight, which enables the vectorization of high-resolution wind stress and surface motion. It also provides a better spectral coverage and therefore a better constraint on the long-wave spectrum.

This presentation will discuss the mapping of ocean wave spectrum into a bi-static SAR spectrum. This work will rely on different and coherent approaches. First, We will present a theoretical analysis extending the historical mono-static closed-form equation (Hasselmann and Hasselmann [1991], Krogstad [1992], Engen and Johnsen [1995]) and relying on bistatic transfer functions and bi-static configuration. This approach allows an easier understanding and interpretable analysis of the bi-static SAR mapping of ocean wave spectra.
This theoretical closed-form equation will be exploited and compared to numerical instrumental simulations which mimics as physically as possible the full observation chain of a prescribed ocean scene. Despite the high computational cost, these simulations offer a much larger panel of possibilities to look at instrumental and sea-state-parameter impacts on the resulting SAR spectra. The bi-static specifications will be emphasized and compared to the mono-static equivalent configuration in order to demonstrate the benefits of Harmony in terms of wave retrieval.

To corroborate the findings of the combined theoretical and numerical analysis, we will rely on existing Sentinel-1 data acquired on the same ocean scene at a slightly different time during consecutive ascending and descending passes. These co-located mono-static acquisitions are not fully equivalent to a multi-static SAR configuration as Harmony, but are representative of, and give insight into, the valuable azimuthal diversity gain to better retrieve the directional properties of ocean wave spectrum.

The three approaches previously presented will show that the additional lines-of sight benefits the retrieval of the wave spectrum. The bi-static companions are sensitive to waves traveling in different directions, which makes the RAR spectral analysis of high interest to the study of wind wave characteristics. The ratios of the intensities vary with direction and wave number and therefore bi-static companions provide new means to help retrieve the directional surface-wave spectrum. The SAR transform is more complex. Still, compared to the mono-static transform the bi-static transform displays improved capabilities, particularly in terms of a larger spectral coverage.

A microwave range transponder has been operating at the CDN1 Cal/Val site on the
mountains of Crete for about 6 years, to calibrate international satellite radar altimeters
in the Ku-band. This transponder is part of the European Space Agency Permanent
Facility for Altimetry Calibration, and has been producing a continuous time series of
range biases for Sentinel-3A, Sentinel-3B, Jason-2, Jason-3 and CryoSat-2 since
2015. As of 18-Dec-2020, the CDN1 transponder has allowed calibration of the new
operational altimeter of Sentinel-6A satellite as it flies in tandem with Jason-3. This
work investigates range biases derived from the long time series of Jason-3 (and
subsequently that of Sentinel-6 since both follow the same orbit) and tries to isolate
systematic and random constituents in the produced calibration results of the
transponder. Systematic components in the dispersion of transponder biases are
identified as of internal origin, coming from irregularities in the transponder instrument
itself and its setting, or of external cause arising from the altimeter, satellite orbit,
Earth’s position in space, geodynamic effects and others. Performance characteristics
of the CDN1 transponder have been examined. Draconic harmonics, principally the 58-
day period, play a significant role in the transponder results and create cyclic trends in
the calibration results. The attitude of the satellite body as it changes for solar panel
orientation contributes an offset of about 7 mm when yaw rotation is off its central
position, and the atmospheric, water mass and non-tidal ocean loadings are
responsible for an annual systematic signal of 10 mm. At the time of writing, all other
constituents of uncertainty seem random in nature and not significantly influential,
although humidity requires further investigation in relation to the final transponder
calibration results.

The Traceable Radiometry Underpinning Terrestrial- and Helio- Studies (TRUTHS) satellite mission is a climate mission led by the UK Space Agency (UKSA) and delivered by the European Space Agency (ESA). One of the main objectives of TRUTHS is to provide in-orbit cross-calibration traceable to SI standards for Earth observation satellite missions. The possibility of in-orbit cross-calibration will enhance the performance of calibrated sensors and allow up to a tenfold improvement in data reliability compared to existing data. The ability to obtain more reliable data will lead to improved future climate models, which are crucial for decision-making and action against climate change.
In the development process of the TRUTHS mission, an End-to-End Mission simulator is used to reproduce different mission configurations and evaluate their performance. As part of this simulator, a scene generator module is required that simulates the TRUTHS Top Of Atmosphere (TOA) radiances for several different land surface types. Reflectance cubes from NASA’s airborne imaging spectroradiometer AVIRIS-NG were used as input data for six different land surface types (ocean, agriculture, forest, snow/ice, clouds and desert). First, an extrapolation of the AVIRIS-NG spectral range to the UV range was performed in dependence of the land surface type. A spatial resampling to 2000 pixel across-track and a spectral resampling to 1nm intervals was necessary to meet the requirements of TRUTHS. Further, a simulated TRUTHS sensor file was generated and used as an input to ATCOR to compute a simulated TOA radiance of TRUTHS. ATCOR is a software product that uses the MODTRAN-5 radiative transfer equation to simulate at-sensor radiances. For the latter step, different solar zenith angles were considered to provide a minimum and maximum solar zenith angle per scene. For validation purposes, a cross-comparison was performed using TOA radiances of AVIRIS-NG and the simulated TRUTHS TOA radiances. The final products are delivered in NetCDF file format and will be used as target scenes to be observed by the TRUTHS sensor and hence to test and evaluate its performance.

IASI radiometric error budget assessment and exploring inter-comparisons
between IASI sounders using acquisitions of the Moon


IASI (Infrared Atmospheric Sounding Interferometer) instruments on-board METOP polar orbiting meteorological satellites are currently used for climate studies [1-3]. IASI-A launched in 2006, displayed 15 years of stable performances and is no longer active. There are still two operational IASI instruments: one on-board METOP-B (launched in 2012) and one on-board METOP-C (launched in 2018). Efforts are continuously being made by CNES to improve IASI data quality during their whole lifetime. For example, the methodology for the spectral calibration was improved for IASI-A and a recent reprocessing was performed by EUMETSAT in order to obtain continuous homogeneous data series for climate studies. Moreover, the on-board processing non-linearity corrections for both IASI-A and IASI-B instruments were improved in 2017, reducing the NedT error in the spectral band B1.
IASI is the reference used by the GSICS (Global Space based Inter-Calibration System) community for inter-comparisons between infrared sounders to improve climate monitoring and weather forecasting. The objective here is to present the errors sources which impact the IASI radiometric error budget considering the uncertainties related to the knowledge of the internal black body (e.g. temperature and emissivity), the non-linearity correction and the scan mirror reflectivity law. This work is performed in the framework of the collaboration with the GSICS community to ensure a stable traceability of infrared sounders radiometric and spectral performances.
Moreover, Moon data are regularly acquired since 2019 by IASI-B and IASI-C to study the possibility to perform absolute and relative calibrations by using these lunar observations. The Moon is often used in the visible domain as a calibration source for satellite instruments, but, until now, it is not the case in the thermal infrared domain. In the framework of this study, a dedicated radiometric model was built to simulate and compare IASI lunar measurements. Inter-comparisons between IASI-B and IASI-C Moon acquisitions showed really promising results with an accuracy of ≤ 0,15K. These results are comparable to the performance of IASI instruments inter-comparisons based on selected homogeneous Earth View spectra.
[1] Marie Bouillon and co, Ten-Year Assessment of IASI Radiance and Temperature, Remote Sens., 12(15), 2393 (2020)
[2] Simon Whitburn and co, Trends in spectrally resolved outgoing longwave radiation from 10 years of satellite measurements, npj Climate and Atmospheric Science 4, article number 48 (2021)
[3] Nadia Smith and co, AIRS, IASI, and CrIS retrieval records at climate scales: an investigation into the propagation of Systematic Uncertainty, Journal of Applied Meteorology and Climatology, vol. 54, issue 7, (2015)

SI-Traceable satellites (SITSat) provide highly accurate data with an unprecedented absolute calibration accuracy robustly tied to international system of units, the SI. This increased accuracy and SI traceability helps to improve the quality and trustworthiness of the measurements performed by the SITSat itself and those of others through in-orbit cross-calibration, enabling the prospect of litigation quality information. Such a system can have direct benefits for the net-zero agenda as it reduces the prospect of ambiguity and debate through the ability to understand and remove biases in a consistent and internationally acceptable manner, hereby creating harmonised interoperable virtual constellations of sensors to support decision-making and monitoring of climate change mitigation strategies accounting for emissions and sinks. The very high accuracy capabilities of SITSats can also provide a benchmark from which change can be monitored so that the intended success of our climate actions can be identified and quantified in as short a time as possible.

In this poster we consider how the ESA Traceable Radiometry Underpinning Terrestrial- and Helio- Studies (TRUTHS) mission can contribute to and investigate some of the benefits of the high calibration accuracy and hyperspectral nature of a mission like TRUTHS in relation to the climate emergency. TRUTHS is a climate mission, led by the UK Space Agency, which is being developed as part of the ESA Earth Watch programme to enable, amongst other things, in-flight calibration of Earth observation (EO) satellites. TRUTHS will establish an SI-Traceable reference in space with an unprecedent calibration accuracy of 0.3% (k=2), over the spectral range of 320 nm to 2400 nm at a ground spatial resolution of up to 50 m.

In particular, we look at how TRUTHS might help to anchor sensor measurements used to estimate sinks and sources of GHG emissions, including ocean and land biology, and track land use changes. We also explore how TRUTHS might help to constrain atmospheric measurements by improving the quality of ancillary information used in the retrievals e.g. aerosols, surface albedo, as well as bias removal in the sensor radiometric gains through in-orbit calibration, enabling harmonised constellations of satellites in support of the stocktake.

For the stocktake, as many GHG monitoring satellites have large field of views and also are anticipated to sit in a variety of orbits, we explore in some detail the impact of solar illumination and view angle on the calibration process. Here we evaluate how TRUTHS can estimate the bidirectional reflectance distribution function (BRDF) of the surface, particularly of typical desert calibration targets, and the impact on uncertainty.

Additionally, regardless of the main purpose of TRUTHS in this context as a ‘metrology laboratory in space’ and calibration reference, we also explore how TRUTHS itself can perform or at least contribute to some of the mitigation-related measurements. Even though the mission is not explicitly designed for many short-term climate action related activities, by virtue of being hyperspectral, of high accuracy and relatively high spatial resolution it can still make a positive contribution and improve spatial and temporal coverage of monitoring. As an example of such measurements, we chose the detection of methane point emitters (e.g. fossil fuel extraction and use facilities, agriculture facilities and landfills), one of the top priorities among the mitigation actions for the next decade.

In summary, the poster will explore the impact and contribution of a SITSat like TRUTHS to the climate action agenda through both direct observations and derived information, improvement of retrieval algorithms and interoperability and accuracy of existing sensors specifically designed for particular variables such as GHG satellites, those monitoring land and ocean biological properties serving as sinks and/or their impact on emissions. The climate emergency requires policy makers and society to have confidence in the information in order to pursue the necessary actions and this needs to be underpinned by data of rigorous and unchallengeable uncertainty.

Establishing an end-to-end uncertainty budget is essentially required for all ECVs of ESA’s Climate Change Initiative (CCI). The reference guide for expressing and propagating uncertainty consists of the GUM and its supplements, which describe multivariate analytic and Monte Carlo methods. The FIDUCEO project demonstrated the application of these methods to the creation of ECV datasets, from Level 0 telemetry to Level 1 radiometry and beyond. But despite this pioneering work, uncertainty propagation for ECVs is challenging. Firstly, many retrieval algorithms do not incorporate the use of quantified uncertainty per datum. Using analytic methods for propagating uncertainty requires completely new algorithmic developments while applying Monte Carlo methods is usually straightforward but leads to proliferation of computational and data curation resources. Secondly, operational radiometry data are usually not associated with a quantified uncertainty per datum and error correlation structures between data are not quantified either. Deriving this information from original sensor telemetry and an according harmonisation with respect to an SI traceable satellite (SITSAT) reference is a future task.

Nevertheless, it is feasible to explore and prepare ECV processing for the use of uncertainty per (Level 1) datum and error correlation structures among data already now, based on instrument specifications and simplifying assumptions. For the Land Cover ECV of the CCI we developed a Monte Carlo surface reflectance pre-processing sequence, which considers three most significant effects: errors in satellite radiometry, errors in aerosol retrieval, and errors in cloud detection. Error correlations between radiometric data are considered using a simplified correlation matrix with a constant correlation coefficient. Such simplified correlation structure can take account of uncorrelated random noise as well as common systematic errors arising, e.g., from radiometric calibration, which affect climate data sets even in the long term, while all other forms of random error average out sooner or later. Errors in aerosol retrieval are considered in a similar way. Errors in cloud detection affect the land cover classification directly. Omission of clouds degrades the accuracy of the ECV dataset whereas false commission reduces its coverage statistics. Our Monte Carlo pre-processing sequence can simulate random and systematic cloud omission and commission errors.

In this contribution, we explain the concept of our Monte Carlo processing sequence and its computational implementation and present proof of concept by verifying the statistical properties of the created surface reflectance ensemble.

The Global Space-based Inter-Calibration System (GSICS) is an initiative of CGMS and WMO, which aims to ensure consistent accuracy among satellite observations worldwide for climate monitoring, weather forecasting, and environmental applications. To achieve this, algorithms have been developed to correct the calibration of various instruments to be consistent with community-defined reference instruments based on a series of inter-comparisons – either directly by the Simultaneous Nadir Overpass (SNO) or Ray-Matching approach – or indirectly using Pseudo Invariant Calibration Targets (PICTs), such as the Moon, desert sites or Deep Convective Cloud as transfer standards. In the former approach contemporary satellites are tied to current state-of-the-art reference instruments, while heritage satellites need to rely on older references. The invariant target approach relies on their characterisation by counterpart reference instruments and is typically applied in the Reflected Solar Band.

The 2020s will see the launch of a new type of satellite instrument, whose calibration will be directly traceable to SI standards on orbit, referred to here as SI-Traceable Satellites (SITSats). Examples include NASA’s CLARREO Pathfinder, ESA’s TRUTHS and FORUM, and the Chinese Space Agency’s LIBRA. The first of these will carry steerable VIS/NIR spectrometers, which will allow corresponding GSICS products to be tied to an absolute scale.

This presentation outlines two approaches being developed to exploit these SITSats. Firstly, by direct comparison of their observations with those of current GSICS reference instruments using Ray-Matching to ensure equivalent viewing conditions over simultaneous, collocated scenes. Secondly, by charactering the current Pseudo Invariant Calibration Targets including Deep Convective Clouds and desert sites, including their BRDF and spectral signature. The challenge of propagating uncertainties through the inter-calibration algorithms to achieve a full traceability chain will be discussed.

To optimise the benefits of such a SITSats requires GSICS to prioritise which reference instruments or PICTs are to be characterised, and close cooperation with the SITSat operators to ensure sufficient acquisitions are available to fully characterise them within the mission lifetime. Ultimately, tying GSICS products to an absolute scale would provide resilience against gaps between reference instruments and drifts in their calibrations outside their overlap periods, and allow construction of robust and harmonized current and historical data records from multiple satellite sources to build Fundamental Climate Data Records, as well as more uniform environmental retrievals in both space and time, thus improving inter-operability.

TRUTHS (Traceable Radiometry Underpinning Terrestrial- and Helio-Studies) is a UKSA-led climate mission that is in development as part of ESA Earth Watch programme, which has the aim of establishing of high accuracy SI-traceability on-orbit to improve estimates of the Earth’s radiation budget and other parameters by up to an order of magnitude. The high-accuracy that its SI-traceable calibration system enables (target uncertainty of 0.3 % (k=2)) allows TRUTHS observations to be used both directly as a climate benchmark and as a reference sensor for upgrading the calibration of other sensors on-orbit.

In order to the assess the proposed instrument design against the strict uncertainty requirements, a rigorous and transparent evidence-based uncertainty analysis is required. This paper describes a metrological analysis of the radiometric processing of the TRUTHS L1b products derived from Top of the atmosphere observed photons, including an analysis of the On-Board Calibration System (OBCS) performance. At the heart of the OBCS is the Cryogenic Solar Absolute Radiometer (CSAR) which provide the primary traceability to SI. The OBCS mirrors concepts used in national standards laboratories for measurement of optical power and spectral radiance and irradiance and in TRUTHS links the calibration of the Hyperspectral Imaging Spectrometer (HIS) to SI.

The analysis follows the framework outlined in the EU H2020 FIDelity and Uncertainty in Climate data records from Earth Observations (FIDUCEO) project, which uses a rigorous GUM-based approach to provide uncertainties for Earth Observation products and contains a number of documentational and visualisation concepts that aid interrogation and interpretation. Initially the measurement functions for each instrument on board TRUTHS are defined, and a corresponding ‘uncertainty tree diagram’ visualisation produced. From this, error effects are identified and described using ‘effects tables’, which document the associated uncertainty, sensitivity coefficients and error-correlation structure, providing the necessary information to propagate the uncertainty to the final product. Combining this uncertainty information allows for the total uncertainty of a quantity (e.g. radiance) to be estimated, at a per-pixel level, which can then be analysed based on the source of the uncertainty or its error-correlation structure (e.g., random, systematic, etc.).

An extension of this analysis is End-to-End Metrological Simulator (E2EMS) for the TRUTHS L1b Products, adding both a forward model of the TRUTHS sensor (input radiance to measured counts), and a calibration model (measured counts to calibrated radiance), to understand the product quality an contributions from the proposed schemes and algorithms.

The CEOS (Committee on Earth Observing System) Cal/Val Portal website: https://calvalportal.ceos.org/ serves as the main online information system for the CEOS Working Group on Calibration and Validation enabling exchanges and information sharing to the wider Earth Observation community within CEOS and beyond.

It provides users with connections to good practices, references and documentations as well as reference data and networks and is a source for reliable, up-to-date and user-friendly information for Cal/Val tasks. The portal facilitates data interoperability and performance assessment through an operational CEOS coordinated and internationally harmonized Cal/Val infrastructure consistent with QA4EO principles.

It is possible to access the various contents within the portal as a guest or by logging-in as a member. As a registered user you gain the rights to view dedicated sections, download/upload documents and contribute to the portal growth (e.g.: access to the document repository, specific datasets, terms and definitions area, etc.). News, announcements or novel content are highlighted on the home page and in the twitter feed (@CEOS_WGCV), providing fresh information from and for the community.

The CEOS WGCV page is the entry point for all the CEOS Working Group on Calibration and Validation (WGCV) subgroups, and hosts the IVOS and the CEOS SAR subgroup websites. The Cal/Val Sites page offers an overview, by a linked tree diagram, of the test sites used for calibration and validation activities. The sites are grouped according to WGCV subgroup domain and applications. The CEOS endorsed sites and the reference networks are distinguished by different colors.

In the Projects section relevant Cal/Val project links are provided subdivided by disciplines: Atmosphere, Land, Cryosphere and Ocean. In the Campaigns section several campaign websites links are provided and categorized, as for projects. The list of Cal/Val software tools and services are presented in the Tools page with corresponding links and descriptions. The portal hosts in the Cal/Val Data section the Modulation Transfer Function (MTF) Reference Dataset – with a dedicated webpage with reference papers and reference imagery, and the Speulderbos forest field database.
The Cal/Val Portal is based on Liferay®, an open source web platform that supports content management and other collaborative tools.

The Fiducial Reference Measurement (FRM), Fundamental Data Record (FDR) and Thematic Data Product (TDP) activities all have a common aim – to provide long-term satellite data products that are linked to a common reference (ideally the SI), with well-understood uncertainty analysis, so that observations are interoperable and coherent. In other words, measurements by different organisations, different instruments and different techniques should be able to be meaningfully combined and compared. These programmes have implemented the principles of the Quality Assurance Framework for Earth Observation (QA4EO), which was adopted in 2008 by the Committee for Earth Observation Satellites (CEOS).

The adoption of QA4EO, and the comprehensive research programme that has followed it, have come from a fruitful and long-term collaboration between scientists working in National Metrology Institutes (NMIs) and the Earth Observation community, and especially the efforts of ESA to embed metrological principles in all its calibration and validation activities.

The European Association for National Metrology Institutes (EURAMET) has recently created the “European Metrology Network (EMN) for Climate and Ocean Observation” to support further engagement of the climate observation and monitoring communities with metrologists at national metrology institutes and to encourage Europe’s metrologists to coordinate their research in response to community needs. The EMN has a scope that covers metrological support for in situ and remote sensing observations of atmosphere, land and ocean ECVs (and related parameters) for climate applications. It is the European contribution to a global effort to further enhance metrological best practice into such observations through targeted research efforts and provides a single point of contact for the observation communities to Europe’s metrologists.
 
In 2020 the EMN carried out a review to identify the metrological challenges related to climate-observation priorities. The results of that review are available on the EMN website (www.euramet.org/climate-ocean) and include 32 identified research topics for metrological institutes. The EMN is now defining a strategic research agenda to respond to those needs. The EMN is also working with the International Bureau of Weights and Measures (BIPM) and the World Meteorological Organization (WMO) to organise a “metrology for climate action workshop” to be held online in October 2022.

Here we present the activities of the EMN and how they relate to the establishment of SI-traceability for satellite Earth Observations.

Society is becoming increasingly dependent on remotely sensed observations of the Earth to assess its health, help manage resources, monitor food security, and inform on climate change. Comprehensive global monitoring is required to support this, necessitating the use of data from the many different available sources. For datasets to be interoperable in this way, measurement biases between them must be reconciled. This is particularly critical when considering the demanding requirements of climate observation – where long time series from multiple satellites are required.

Typically, this is achieved by on-orbit calibration against common reference sites and/or other satellites, however, there often remain challenges when interpreting such results. In particular, the degree of confidence in the resultant uncertainties and their traceability to SI is not always adequate or transparent. The next generation of satellites, where high-accuracy on-board SI-traceability is embedded into the design, so-called SITSats, can therefore help to address this issue by becoming “gold standard” calibration references. This includes the ESA TRUTHS (Traceable Radiometry Underpinning Terrestrial- and Helio- Studies) mission, which will make hyperspectral observations from visible to short wave infrared with a target uncertainty of 0.3 % (k = 2).

To date, uncertainty budgets associated with intercalibration have been dominated by the uncertainty of the reference sensor. However, the unprecedented high accuracy that will be achieved by TRUTHS, and other SITSats, means that the reference sensor will no longer be the dominant source of uncertainty. The accuracy of cross-calibration will instead be ultimately limited by the inability to correct for differences between the sensor observations in comparison, e.g., spectral response, viewing geometry differences. The work presented here aims to assess the impact of these differences on the accuracy of intercalibration achievable using TRUTHS as a ‘reference sensor’ and evaluate to what extent these can be limited through appropriate design specifications on the mission.

A series of detailed sensitivity analyses have been performed to evaluate how the intercalibration uncertainty for TRUTHS and a given target sensor can be best minimised, based on potential TRUTHS design specifications. This includes the impact of TRUTHS’ bandwidth and spectral sampling definition, which was studied using a radiative transfer model to investigate how well TRUTHS can reconstruct target sensor bands for comparison. A similar simulation-based approach is used to evaluate the sensitivity of intercalibration to TRUTHS’ spatial resolution. Target sensor (Sentinel-2 MSI) images are resampled as a proxy to simulate TRUTHS images at a range of spatial resolutions. The ability of TRUTHS to reconstruct target sensor images is then assessed by resampling the simulated-TRUTHS image back to the spatial resolution of the target sensor. These simulation studies were carried out for a range of sites, including CEOS desert pseudo-invariant calibration site (PICS) Libya-4, representing the types of scenes that are used as targets in the sensor intercalibration process. Target sensors simulated in these studies include the widely used sensors Sentinel-2 MSI, Sentinel-3 OLCI and Suomi-NPP VIIRS, as they are representative of many of the types of sensors TRUTHS will be used to calibrate.

The main objective of the project “Precise Orbit Determination of the Spire Satellite Constellation for Geodetic, Geophysical, and Ionospheric Applications” (ID no. 66978), which was approved on 7 September 2021 in the frame of an ESA Announcement of Opportunity (AO), is to generate and validate precise reference orbits for selected Spire satellites and, based on this, to ingest and assess the requested Spire GPS data into three scientific applications, namely gravity field determination, reference frame computations, and ionosphere modelling to study the added value of the Spire GPS data. Due to the fact that the Spire constellation populates for the first time the Low Earth Orbit (LEO) layer at different inclinations with a large number of satellites, which are all equipped with high-quality dual-frequency GPS receivers, it opens the door to significantly strengthen all of the three above mentioned scientific applications.

In the initial phase of the project the focus will be on the precise orbit determination (POD) of selected Spire satellites. Two independent, state-of-the art software packages, namely the Bernese GNSS Software and ESA’s NAPEOS software, will be used for this purpose. This will allow for inter-comparisons, a role model that is inherited from the work of the POD Quality Working Group of the Copernicus POD service. It will enable an independent quality and integrity assessment of the Spire inputs and products.

We will analyse the quality of the Spire GPS code and carrier phase date and validate antenna phase centre calibrations. Based on this we will determine reduced-dynamic and kinematic orbits for selected Spire satellites. Eventually we will evaluate the quality of the reconstructed orbits by means of orbit overlap analyses, cross-comparisons of kinematic and reduced-dynamic orbits computed within one and the same software, and cross-comparisons of the orbits derived with the Bernese GNSS Software and ESA’s NAPEOS software, as well as comparisons to the orbits provided by Spire.

The present paper aims to showcase the portfolio of R&D activities that in these years the Italian Space Agency (ASI) is carrying out in collaboration with the national research community, to address scientific applications based on the exploitation of Synthetic Aperture Radar (SAR) data from the national mission COSMO-SkyMed, as well as Copernicus and other bilateral cooperation missions (e.g. SAOCOM).
The focus is on algorithms development and integration of multi-mission SAR data that are collected at different wavelengths, in light of the current unprecedentedly wide spectrum of observations of the Earth’s surface ranging from X- to L-band provided by SAR missions in Europe and beyond.
Within such a framework, COSMO-SkyMed First Generation, TerraSAR-X, Sentinel-1 and ALOS-2 satellites have been operating since several years. On the other side, SAOCOM, NovaSAR, COSMO-SkyMed Second Generation and RADARSAT Constellation Mission satellites have been successfully launched starting from late 2018. Therefore, the geoscience and remote sensing community is increasingly provided with: (i) continuity of observations with respect to previous SAR missions; (ii) opportunities to task collection of spatially and temporally co-located datasets in different bands.
The challenge is now to develop processing algorithms that can make the best out of this multi-frequency observation capability, in order to address a multitude of scientific questions and downstream applications. These applications include, but are not limited to: retrieval of geophysical parameters; land cover classification; interferometric (InSAR) analysis of ground deformation and for structural health monitoring; generation of value added products useful to end-users and stakeholders, for example for civil protection, disaster risk reduction, resilience building, sustainable use of natural resources.
In the framework of the ASI’s roadmap towards the development of SAR-based scientific downstream applications, recent R&D projects have act as the foundational steps to define, develop and test new algorithms for SAR data processing and integration. The R&D activities have intentionally focused from the statement of the initial scientific idea (Scientific Readiness Level – SRL 1, according to ESA EOP-SM/2776 scale) to, at least, the demonstration of the proof of concept (SRL 4) through extensive analyses by means of dedicated experiments and ground-truth validation.
Building upon this heritage and in order to move forwards (also in terms of higher SRL), ASI has recently launched a dedicated programme named “Multi-mission and multi-frequency SAR”. It supports R&D projects proposed by leading experts in the fields from national public research bodies and industry – also in the framework of international partnerships – to design, develop and test innovative methods, techniques and algorithms for exploitation of multi-mission/multi-frequency SAR data, with credible perspectives of engineering and pre-operational development, thus being able to contribute to the improvement of socio-economic benefits of the end-user community.
The current projects address the following R&D areas of specific interest: agriculture, urban areas, natural hazards, cryosphere, sea and coast; alongside the common cross-cutting topic of validation of products generated from multi-frequency SAR data by using ground-truth data (Figure 1).
In light of the experiences gained during recent R&D projects and at nearly one year since the initiation of the multi-mission and multi-frequency SAR projects, the present talk will outline the novelty of the methodological approaches under testing and demonstration, ongoing activities and results achieved, with a particular focus on the integration of Sentinel-1, COSMO-SkyMed, SAOCOM and ALOS-2 data.
Discussion will include:
- Lessons learnt about the role played by regularly acquired multi-frequency and multi-mission SAR time series (also in combination with other EO data) for observation continuity and investigation of long-term processes, either natural and anthropogenic ones;
- Benefits and limitations of acquisition programmes over the national territory and across different locations in the world vs. user requirements;
- The added value brought by the polarimetric SAR capability, e.g. for retrieval approaches of geophysical parameters;
- The importance of coupling satellite observations with instrumented sites and contextual surveys, for both calibration/validation activities and integrated analyses;
- Reflections about the perspectives towards future pre-operational implementation in scientific downstream applications.

The paper describes the realization of an access point to the CONAE SAOCOM mission, enabling the ordering and the dissemination of the products acquired by the SAR sensor over the geographic area in which the Italian Space Agency (ASI) has exclusive rights of exploitation of the data. SAOCOM (Satélite Argentino de Observación COn Microondas) is a L-band SAR based remote sensing constellation owned by the Argentinean Space Agency CONAE (Comisión Nacional de Actividades Espaciales) and formed by two identical 1A and 1B satellites, launched in 8 October 2018 and 30 August 2020 respectively. Within the collaborative project named SIASGE (Italian-Argentinian satellite system for Disaster Management and economic development), a certain amount of SAOCOM data acquisition and processing resources has been reserved to ASI for exclusive use in the so-called Zone of Exclusivity (ZoE) placed in the 10W-50E longitude range and 30-80N latitude range. In such zone ASI has the right to use the SAOCOM system freely, fully and up to the saturation of the granted resources (around 150s of sensing time per orbit), for scientific and institutional purposes by users constituted by the agency internal personnel or by people who -for the strict scopes of SAOCOM mission exploitation- became affiliated to ASI. At the time of writing, the ASI SAOCOM access point offers the capability to select and download products over the ZoE chosen in an archive containing more than 6k images but also experimental services for ordering the processing of SAOCOM data at various levels (from complex slant range up to geocoded) and programming new acquisition in the ZoE.
The development of the access point to mission resources has been based on the following approach and concepts:
• Reuse a reliable archive/catalogue system which is well known and full proven by the Remote Sensing community, possibly released under an Open Source license
• Maintain simplest possible interfaces for registration of users and for requesting products and new acquisitions
• Develop the access point in an incremental way, enriching the basic functions like registration and product dissemination with upper capabilities as managing new acquisitions request
• Use a storage/computation informatic infrastructure based on cloud resources owned by public Italian organizations
• Maintain a strict cooperation with the CONAE SAOCOM team for improving the interactions (at interfaces, archive contents, etc level) between the ASI access point and the Argentinean mission GS
Under these rationales and concepts, the access point has been realized with:
• the ESA developed DHuS - Data Hub System as the catalogue / archive system, widely adopted in the Copernicus Sentinel GS, which offers both a traditional web based human interface but also OpenSearch and OData M2M (Machine to Machine) product search and download capabilities
• ASI internally developed SW for handling user registration, incremental archive filling (through discovery/download actions over the Argentinean mission GS) and for product reprocessing and new acquisition programming experimental functions
• A cloud infrastructure based on OpenStack framework running on GARR, the Italian ultra-broadband network dedicated to the education and research community, having as main objective to provide high-performance connectivity and to develop innovative services for the daily activities of teachers, researchers and students and for international collaboration
• The collaborative support of CONAE for the transfer of the entire SAOCOM product archive on the ASI ZoE (near 130k product) and for the set-up of M2M interfaces by the SAOCOM GS and the ASI access point

The Advanced Land Observing Satellite (ALOS) was launched by the Japan Aerospace Exploration Agency (JAXA) in January 2006. It carried three sensors: the Advanced Visible and Near Infrared Radiometer type 2 (AVNIR-2), the Panchromatic Remote-sensing Instrument for Stereo Mapping (PRISM), and the Phased Array type L-band Synthetic Aperture Radar (PALSAR).

The data for observations over Africa, the Arctic and Europe were collected by two European Space Agency (ESA) ground stations as part of the ALOS Data European Node (ADEN), under a distribution agreement with JAXA, and subject to a recent bulk processing campaign. The latter focused on processing data from the AVNIR-2 and PRISM sensor only, from Level 0 (raw) – Level 1C (orthorectified).

The quality control activities concerning the L1B1 and L1C datasets for both sensors were performed prior to their release / dissemination to users. The quality control activities concerning the brand new L1C products, which were generated using an instrument processing facility developed by the German Aerospace Centre (DLR), included DIMAP product format (including the ESA EO-SIP product wrapper format), geometric and radiometric calibration quality checks. The results of these quality checks, which will be presented in more detail in the poster, generally indicate the data quality is nominal.

----

The Landsat programme, jointly operated by the United States Geological Survey (USGS) and the National Aeronautics and Space Administration (NASA), provides the world’s longest running system of satellites for the medium-resolution optical remote sensing of land, coastal areas and shallow waters.

The data acquired over Europe by the European Space Agency (ESA), using their ground stations (in co-operation with the USGS and NASA), has been subjected to a recent bulk L1C reprocessing campaign. The reprocessed dataset, generated by the Systematic Landsat Archive Processor (SLAP) Instrument Processing Facility (IPF), developed by Exprivia for the Landsat 5 Thematic Mapper (TM) and Landsat 7 Enhanced Thematic Mapper Plus (ETM+) datasets, allows for the historical Landsat products to be updated / aligned with the highest quality standards that can be achieved with the current knowledge of the instrument (e.g. geometric processing via applications of orbit state vector files, an updated TM/ETM ground control points database and digital elevation models).

The quality control activities concerning the L1C datasets for both sensors were performed prior to their release / dissemination to users. The quality control activities concerning the new L1C products, included product format (including the ESA EO-SIP product wrapper format), geometric and radiometric calibration quality checks. The results of these quality checks, which will be presented in more detail in the poster, generally indicate the data quality is nominal.

SAOCOM-1A and -1B are a pair of satellites developed by the Comisión National de Actividades Espaciales (CONAE) of Argentina launched in October 2018 and August 2020, respectively, orbiting on a Sun-synchronous orbit at 620 km altitude 180 deg. apart, and providing imaging of the Earth's surface with an effective revisit time of 8 days. Their main payload is a full-polarimetric Synthetic Aperture Radar (SAR) operating in L-band with selectable beams and imaging modes.
This paper reports the first result of collaborative work between CONAE and CSL, investigating the use of polarimetric and interferometric SAR signatures to detect changes in agricultural zones in Argentina. This work is the natural continuation of a previous pre-SAOCOM activity [1] that made use of airborne SAR images from the national SARAT sensor and the NASA/JPL UAVSAR, and from the spaceborne JAXA ALOS-PALSAR-2, all operating in L-band.
The test site of interest is referred to as the SAOCOM Core Site, a highly agricultural zone within the Pampas region, located in the surroundings of Monte Buey (a small rural village southeast of the Córdoba province, Argentina (-32º 55', -62° 27')). This site is regularly imaged by the SAOCOM satellites, and regular field measurements are carried out in conjunction with image acquisitions.
We have used full-polarimetric, Stripmap-mode SAOCOM-1A images over the region of interest, acquired between March 2019 and February 2020 in both right-looking ascending and descending orbits, making a temporal series of useable interferometric pairs covering a full year.
Each image was the subject of polarimetric processing, involving generation of backscattering coefficient and Radar Vegetation Index (RVI) maps, Pauli decomposition of the scattering vector, generation and diagonalization of the polarimetric coherence matrix, resulting in derived quantities like the entropy, the anisotropy, and the alpha angle. Interferograms and coherence maps were generated by InSAR processing of the interferometric pairs. Finally, adding full polarization information to the interferometric, i.e., carrying out polarimetric interferometry (PolInSAR) processing, optimized coherence maps were generated, allowing to highlight one or the other backscatter mechanism and follow the evolution along a full season.
This multi-dimensional information was put in relation to terrain events by cross-correlation with field data. All the products were co-registered in order to perform time-series analyses for change detection.
This work was performed under a Belgium-Argentina bilateral collaboration. The CSL contribution was supported by the Belgian Science Policy Office.
Reference
[1] Danilo J. Dadamia, Marc Thibeault, Matias Palomeque, Christian Barbier, Murielle Kirkove and Malcolm W.J. Davidson, “Change Detection Using Interferometric and Polarimetric Signatures in Argentina, 8th International Workshop on Science and Applications of SAR Polarimetry and Polarimetric Interferometry”, ESRIN, Frascati, 23-Jan-2017.

Monitoring transportation for planning, management, and security purposes in urban areas has been growing in interest and application by various stakeholders. Since the late 1990s, commercial very-high-resolution (VHR) satellites have been used for developing vehicle detection methods, a domain previously governed by aerial photography due to superior spatial resolution. Despite the apparent advantages of using air- or drone-borne systems for vehicle detection, several methods were introduced in the last two decades, utilizing space-borne VHR imagery (e.g., QuickBird, WorldView-2/3) with meter (multispectral bands) to submeter (panchromatic band) resolutions. Several of the applications applied machine learning for identifying parking cars. However, for detecting moving cars, two sensor capabilities have been utilized: (1) stereo mode by either satellite constellation or by body pointing abilities; and (2) a gap in the acquisition time between the push-broom detector sub-arrays. Changes in the location of moving objects can be observed between image pairs or across spectral bands, respectively. Both cases require overcoming differences in ground sampling distance and/or prerequisite spectral analyses to identify suitable bands for change detection.
Since January 2018, new multispectral products have been available for the scientific community provided by the Vegetation and Environmental New Micro Spacecraft (VENµS). This mission is a joint venture of the Israeli Space Agency (ISA) and the French Centre National d’Etudes Spatiales (CNES). The overall aim of the VENµS scientific mission is to acquire frequent, high-resolution, multispectral images of pre-selected sites of interest worldwide. Therefore, the system is characterized by the spatial resolution of 5 m per pixel (the upcoming mission phase will increase the spatial resolution to 4 m per pixel), the spectral resolution of 12 narrow bands in the visible-near infrared regions of the spectrum, and revisit time of 2 days in the same viewing and azimuth angles.
Here we demonstrate the VENµS capability to detect moving vehicles in a single pass with a relatively low spatial resolution. The VENµS Super Spectral Camera (VSSC) has a unique stereoscopic capability since two spectral bands (number 5 and 6), with the same central wavelength and width (620 nm and 40 nm, respectively), are positioned at extreme ends of the focal plane (Figure 1). This design enables a 2.7-sec difference in observation time. We took a straightforward approach to create a simple spectral index for moving vehicles detection (MVI) using these bands. Since the two bands are identical, there is no need for prior image analyses for dimensionality reduction or geometric corrections, as required for other sensors. Each moving vehicle is represented by a pair of bright blob-shaped clouds of pixels on a darker background (Figure 2). The center of each cloud in the pair is determined based on the same methodology used to identify the barycenter of a multi-particle system, where the MVI values replace the masses of the particles. Once the center of each cloud is known, the velocity vector, i.e., speed magnitude and orientation, can be extracted by geometrical considerations.
Results show successful detection of moving small- to medium-size vehicles. Especially interesting is the detection of private cars that are on average 2-3 m smaller than the ground sampling distance of VENµS. We effectively detected vehicle movement in different backgrounds/environments, i.e., on asphalt and unpaved roads, as well as over bare soil and plowed fields, and at different speeds, e.g., 61 km/h for a car over an asphalt road, and 19 km/h for vehicles on unpaved road. A speed of 111 km/h was calculated for a heavy train. This speed is in line with the engine speed limit and the regulations applied by Israeli authorities, providing estimation for MVI accuracy.
The MVI benefits from the coupling of a unique detector arrangement of the Super Spectral Camera onboard VENµS. In addition, the very high temporal resolution of 2 days makes VENµS products an attractive input for vehicle detection applications, particularly for operations that require monitoring on a nearly daily basis. It appears to be cost-effective compared to VHR commercial satellite and complex UAV-base monitoring systems. Furthermore, the MVI suggests that such bands arrangement is highly effective and should be considered for future space missions, primarily for surveillance and transportation monitoring.

Passive microwave observations in L-band are unique measurements that allow a wide range of applications, which in most cases cannot be done at other wavelengths: accurate absolute estimations of soil moisture for hydrology, agriculture or food security applications, ocean salinity measurements, detection and characterization of thin ice sheets over the ocean, detection of frozen soils, monitoring above-ground biomass (AGB) to study its temporal evolution and global carbon stocks, measurements of high winds over the ocean… The Soil Moisture and Ocean Salinity (SMOS) satellite, launched by ESA in 2009, which has performed for the first time systematic passive observations at L-band, has allowed to discover some of the previous applications. SMOS L-band data play a central role in the ESA Climate Change Initiative (CCI) for Soil Moisture and Ocean Salinity. Passive L-band data also contribute to the CCI Biomass. This European mission has been followed by two other L-Band missions from NASA: Aquarius and SMAP (Soil Moisture Active Passive).

In the last years, scientific and operational users were requested to contribute to a survey of requirements for a future L-band mission. One of the outcomes of this survey is that most of the applications require a resolution of around 10 km. This is the objective of SMOS-HR (High Resolution) project, a second generation of SMOS mission: the continuation of L-band measurements with an unprecedented native resolution improving by a factor of 2 to 3 that of the current generation of radiometers such as SMOS and SMAP.
In this paper, we will present this SMOS-HR project which is currently under study at CNES (the French Centre National d’Etudes Spatiales) in collaboration with CESBIO (Centre d’Etudes Spatiales de la BIOsphère) and ADS Toulouse (Airbus Defence & Space) which has been contracted by CNES for the instrument definition.

The main challenge for this CNES study is to find the best trade-off to satisfy most needs with “reasonable” mission requirements (i.e., feasible for an acceptable cost). The core mission objective for SMOS-HR is to increase the spatial resolution at least by a factor of two with respect to SMOS (< 15km at Nadir) while keeping or improving its radiometric sensitivity (~0.5-1 K) and with a revisit time no longer than 3days. Taking into account the mission and system level requirements, a new definition of an interferometric microwave imaging radiometer has been studied.
The first step has been to select the antenna array configuration: cross-shaped arrays, square-shaped arrays (which imply a Cartesian gridding), Y-shaped arrays and hexagon-shaped arrays (which imply a hexagonal gridding) have been compared. A cross shape has been selected as the best option because it allows to reduce the aliasing in the reconstructed images by adequately choosing the position of the elementary antennas along the four arms and because its accommodation is simpler than for some other configurations. The result is an instrument with 171 elementary antennas regularly spaced along the arms (~1lambda) and an antenna with an overall size of ~17 meters tip-to-tip. Then, the optimal concept for the SMOS-HR instrument consists of a hub located on the platform, carrying a dozen central antennas, and four deployable arms attached to the platform, carrying about 40 antennas each. The SMOS-HR hub gathers a Central Correlator Unit computing the correlations for all antenna pairs and generating a clock signal for instrument synchronization. The feasibility of on-board processing for Radio-Frequency Interference (RFI) mitigation has also been addressed to overcome the limitations faced on SMOS with on-ground processing. Adding this function for SMOS-HR represents another major improvement compared to SMOS (on top of the resolution improvement).
As a risk reduction activity, a breadboard of a part of the Central Correlation Unit is defined, developed and tested by ADS during this study in order to assess the achievable performances and functionalities of SMOS-HR on-board processing.
Finally, the SMOS-HR phase A has also been the opportunity to explore innovative calibration strategies based on SMOS lessons learnt.

As a synthesis, this talk will present successively:
• SMOS-HR mission and system level requirements,
• The main trade-off at instrument and sub-systems levels (antenna configuration, deployment structure, elementary antenna, on-board correlator, RF receiver, power and local oscillator distribution, calibration strategy…),
• The current results of the correlator breadboard pre-development.

LibGEO is a multi-sensor geometric modeling library, with high location precision. The library is designed to be used at different steps of an Earth Observation mission: prototyping, ground segments, calibration. It is first developed to meet CNES/Airbus Defense and Space, CO3D mission requirements and then to be the CNES reference library for geometry.

The base function of LibGEO is the direct location which returns ground coordinates for each pixel coordinate of the image. It supports both mathematical (grid or RPC) and physical modeling. For physical modeling, a line of sight is built from the detector and is transformed using, for example, rotation, translation, homothety, mirror reflection, to get the line of sight in the International Terrestrial Reference Frame (ITRF). With the position of the platform and an ellipsoid model of the earth, the ground position can be computed. Other location functions are provided such as inverse location, intersection on DEM, colocation. Each location function has a grid implementation, which creates grids to resample images in different geometries (including orthoimages). LibGEO also deals with stereo images for 3D model reconstruction. It has the ability to intersect lines of sight from correlated image points to get a 3D point. LibGEO computes the epipolar geometry that allows dense correlation of the 3D reconstruction.

Native location is not precise enough for some applications, therefore an optimization of the model parameters can be done by using Ground Control Point (GCP) and tie points in image geometry. The improvement of absolute and relative location is done where the user can set uncertainties on both observations and model parameters. During the optimization process, points with higher residual errors are filtered out using statistical methods.

Supported sensors are Pleiades HR, Sentinel-2 MSI (L1B), CO3D for now and some will be added soon: Thrisna, 3MI/Metop SG, Microcarb. The library is built to be generic, other sensors could easily be supported by simply plugging in a product format handler.

The library is designed to be easily integrated in any operational processing chain thanks to its C++ API but it is also user friendly for prototyping and expertise through the Python API.

The Government of Canada (GC) uses multiple sources of data to provide services to Canadians. Given the geographic size of Canada and the need for data collection beyond our landmass, this can often be most efficiently accomplished through Earth Observation. The most critical of these are the RADARSAT series of satellites. Using a powerful synthetic aperture radar (SAR) to collect digital imagery, the RADARSAT series of satellites can “see” the Earth day or night and in any weather condition. The next generation solution is currently being investigated under the Earth Observation Service Continuity (EOSC) program. Due to the wide range of user’s needs, the EOSC initiative considers a broad source of input data including free and open data, commercial purchase of data, international cooperation and a dedicated SAR system. This paper will look at some initial analysis on undertaken under EOSC.

As a first step, a list of User Needs has been collected from various Canadian Federal User Departments. The list of user needs has been consolidate under the Harmonized User Need document. A few key consideration could be extracted from this list of User Needs. The area and the coverage frequency increased compared to the RCM requirements and capabilities. Established applications such as ice monitoring would benefits from both an increase in the coverage frequency and resolution. Even for these established applications, gaps remains to measure some parameters of high interest such as ice thickness. Finally, the document highlight the importance of the access to multi-frequency data for several needs.

The second step was to perform a series of option analysis studies with eight industrial partners to ensure a wide coverage of the potential solution that could satisfied the complex set of user needs. Although no specific solution has been selected at this stage, the studies generally pointed to some extend toward some similar findings. A dedicated C-Band resource will be needed to meet the User Needs including some form of multi-aperture/digital beamforming capability will be required to meet the swath and resolution requirements. Challenges remains to simultaneously meet all User Needs as a broad range of frequency, polarization, coverage and resolution is required often conflicting over similar or adjacent AOI. Free and open, commercial data and international cooperation with other existing system are key to respond to these User Needs while limiting the overall system complexity and cost.

Targeted technology development activities are being to address item of lower technology readiness including enabling technologies for multi-aperture/digital beamforming antenna but also ground segment technology development to provide a better integrated planning of the dedicated EOSC resources taking into consideration all available external sources of data.

Current scenario is moving towards the implementation of tools able to comply with application needs, on-board: to have the information required by end-users at the right time and in the right place. And the place is more and more often becoming the space segment, where the availability of actionable information can be a game-changer.
In this approach part of the EO value chain is transforming. Value is shifting from the sensed data (that nowadays are becoming a commodity) to “insights” and actionable information. So, components of the chain are being moved from user’s desktop to the cloud and from ground to space. Actually, as a final result, user will no longer need to be aware of what data are providing the information he looks for, or where these are stored and processed. The application will be the core and the details related to its workflow (data acquisition, processing, selection, information extraction…) can even be completely transparent to users. Or practically user may only define what he really cares of, everything else
This is the scenario the AI-eXpress services (AIX in short) are enabling. AIX makes available satellite resources and on-board applications as-a-service. Customers can pick-up the application they need from the AIX app store, configure and run it on the satellite already in orbit. The system will take care of scheduling the data acquisition, transforming data into actionable information and also raising near real-time alarms when services require. It is based on the SpaceedgeTM on-board artificial intelligence-based application framework, on distributed ledger technologies (blokchain) machine-to-machine interfaces, on the high-performance computing cluster and finally on the ION cargo spacecraft vehicle.
AIX is a gamechanger. It processes data where it’s more convenient, starting on-board at the “space edge”; it turns EO product generation into services, making the satellite transparent; it makes on-board resources flexible enough to fit to different applications and address different needs, thanks to AI and DLT technologies advances.
AIX fosters the transition from a traditional space model to a really commercial one, reducing bottlenecks and barriers, enabling new market opportunities to flourish and enhancing the effectiveness of the services delivered to the ground.
Emerging NewSpace companies may now test their innovative AI algorithms and their proof-of-concept directly in space and prove their value to the market. Traditional space institutions and research entities may test a new approach changing from “makers” to “enablers”.
AIX builds an infrastructure open to integrate third-party resources and services and aims at building a full eco-system. Thanks to the quick service deployment test and operational capabilities, it candidates to be a strategic assets for commercial applications ranging from oil and gas asset monitoring and management to energy networks (market estimated in 1.3 B€ in 2029 by NSR).
It also enables a large variety of government services supporting both the ESA “accelerators” strategy and the main pillars of the EU Green Deal and the Digital Strategy and fits, as well, as a Copernicus contributing asset in line with the last Request for Ideas for new Copernicus Contributing Missions.

As the advent of New Space becomes reality – new satellite tasking strategies, increased acquisition capacities, EO data distribution channels and user expectations are all changing out of recognition.

New Space affects the traditional EO operational scenario, which relies currently on strict boundaries between data and value added or data stream service providers, however it brings to the CCM activity many disruptive innovation approaches and the promise of new solutions to complex existing challenges, as e.g. fast response times combined with smooth data delivery. Cloudification of workflow processes improves greatly the product availability for users in terms of usability; the data is immediately accessible and exploitation can happen directly on cloud platforms, minimizing product dissemination flows and as well the time-to-exploitation costs. Data as a Service (DaaS) is a consolidated approach, and evolving EO data marketplaces are offering domain specific tool support and downstream applications to maximise the take up and utility of the data by the Copernicus user community. Collaboration and chaining of products to tailor specific user requirements will be a challenge to maximize the exploitation and re-use of EO data.

At the same time, we have new questions that directly impact service sustainability. Among these is how to collaboratively build and promote best operational practices among the growing number and diversity of emerging, new, established actors. The flexible management of demand-oriented data offer and the improvement of operational processes in terms of standardization and simplification are big new challenges for connecting the Copernicus user community quickly to the necessary source data. New Space is changing the landscape of CCM providers, and requires managing operational scenarios in which scalability and diversity are the main drivers, and an Ecosystem of related services concentrating on streamlining and simplification is paramount. Within such an Ecosystem clear roles and responsibilities need to be guaranteed as the reliance on independence and brokerage becomes a key component. In this new paradigm new actors can enter the scenery as the discovery/gatekeeper role, aiming to understand trends in new space technologies, liaising with the Service users to foresee and anticipate coming needs and implement these within the service.

The NovaSAR mission is a UK technology demonstration mission of a small Synthetic Aperture Radar (SAR) satellite. It is a partnership between the UK Space Agency (UKSA), Surrey Satellite Technology Limited (SSTL), the Satellite Applications Catapult (Catapult) and Airbus, with UKSA, SSTL, the Commonwealth Scientific and Industrial Research Organisation (CSIRO), the Indian Space Research Organisation (ISRO) and the Department of Science and Technology-Advanced Science and Technology Institute (DOST-ASTI) sharing the observational capacity of the satellite. NovaSAR-1 was launched in September 2018, with the start of the service starting in late 2019, with a nominal lifespan of 7 years.
NovaSAR was built as a low-cost payload demonstrator, with a manufacturing cost of just 20% of traditional SAR satellites, while ensuring flexible and good performance SAR imaging capabilities. It is a S-band SAR satellite, opening up new observation capabilities to users since most space-borne SAR missions have so far focussed on C-, L- or X-band. It has a revisit time of around 14 days, and a range of acquisition modes available. More importantly, it is the first civilian SAR mission to carry an AIS receiver on board, meaning simultaneity of SAR observation and AIS message reception which has not been possible before. To strengthen this capability, it is equipped with a maritime acquisition mode, which is designed to maximise ship detection over large areas of sea or ocean.
Analysis Ready Data (ARD) has become an increasingly important feature of making satellite mission data more accessible and usable by a wider audience, including non-specialists in Earth Observation. An ever-increasing number of services and applications rely on ingesting and analysing ARD, and so making NovaSAR data available in an ARD format is seen as key for all mission partners to realise their key mission objectives. Among these are to increase the uptake of medium-resolution SAR data through the development of novel applications, supporting respective government-mandated scientific objectives and increasing national expertise in the use of SAR data.
This study will show the creation of a NovaSAR ARD pipeline, with a large collaborative effort from mission partners to align ARD processing flows and work together to resolve some ongoing issues with the NovaSAR data. The aim is to produce ARD alongside level 1 NovaSAR data, that are compliant with CEOS ARD For Land (CARD4L) standards, for both Stripmap and ScanSAR acquisition modes, thus covering all NovaSAR acquisition modes (with the exception of maritime mode). A timeline towards the goal of routinely making available NovaSAR ARD, along with details of specific applications this will enable, will be presented.

As the global constellation of satellites and missions grows, so access to this constellation becomes more complex and challenging for large organisations and institutions. Many such organisations have perennial needs for imagery but require the variety of image sources available to cover their wide variety of use cases. While some satellites provided the best possible spatial resolution, others provide very high temporal resolution, while others providing key imaging bands. These image sources are complementary and all part of a necessary solution for large organisations that wish to have robust and flexible access to the constellation as a whole. The challenge with this is that simply procuring imagery from these various sources through independent and separately negotiated contracts does not provide a convenient solution for these organisations. It is cumbersome, inefficient and inflexible. As well as dealing with multiple operational, ordering and delivery interfaces, there are commercial challenges around pricing, licensing terms and supplier service level performance. Consolidation is required in order to present the customer with a workable and efficient multi-supplier solution to their imaging needs. Consolidation takes makes many forms, but the following aspects are key: a single enterprise platform that specifies the terms for supplier on-boarding and compliance and acts as a vehicle for supplier lifecycle management and user accounts and budgets on a project-by-project basis; an operational dashboard for requesting and delivering imagery from on-boarded suppliers as well as hosting and archive management; various technical support tools (image alerts from aggregated catalogues, multi-mission planning for assisted tasking), and standardisation as far as possible (mandatory licensing terms, pricing for particular image configurations, etc). The enterprise platform can be evolved over time. New suppliers can be onboarded over time and terms and standards can be upgraded as appropriate. In terms of the operational dashboard, this can include access to aggregated catalogues from the on-boarded suppliers (i.e. our EarthImages platform) and standardised tasking requests which are carried out either in a managed or competitive manner by the suppliers, depending on their ability to meet the specification of the request (i.e. Earthimages-on-Demand, developed under funding from ESA). We will describe this new platform and show how it could play a role in the Copernicus CCM programme.

GNSS Radio Occultation (RO) observations from space have been successfully demonstrated since many years and their value for weather prediction is indisputable. The demand for higher spatial and temporal RO observations is steadily increasing as more and more applications in the downstream market rely on pin point accurate weather prediction. The exploitation of radio waves not only bended within the atmosphere but also reflected on the earth surface for analysis of surface features is a more recent development and from the perspective of the instrument architecture closely related to the RO method.
The GRAS instrument on MetOp first generation still provides observation data of unprecedented quality which is a result of high end GNSS receiver design including antenna, clock, LO and RF design. In contrast to this the PRETTY mission will provide GNSS passive reflectometer (PR) data with simple RF and antenna architecture which is suitable to be accommodated on a 3U CubeSat. With this approach good performance can be reached at a fraction of the cost. For the PRETTY development a COTS approach has been followed, and the costs for the signal processing and development of the high-level software have been significantly reduced.
We are presenting a concept in which the GRAS instrument is enhanced with the PRETTY signal processing part allowing to provide RO and PR observations to be performed with the same high-end performance. The high gain antennas and the front end based on the Saphyrion G3 architecture are used to provide baseband samples to the two different signal processing cores, one based on the AGGA-4 architecture the other one on a System on Chip using an ARM processor. The enhanced GRAS instrument can be used as hosted payload on small satellites and will provide both high quality RO and PR observations.

For the planned NanoMagSat constellation mission the University of Oslo (UiO) will contribute a multi-needle Langmuir probe (m-NLP) system. The m-NLP is a compact, light-weight, power-frugal instrument providing in situ ionospheric plasma density measurements. Typically, Langmuir probes operate by sweeping through a range of bias voltages in order to derive the plasma density, a process that takes time and hence limits the temporal resolution to a few Hz. However, the m-NLP operates with fixed bias voltages, such that the plasma density can be sampled at 2kHz, providing a spatial resolution finer than the ion gyroradius at orbital speeds. The NanoMagSat m-NLP design is based on heritage from sounding rockets, CubeSats, SmallSats, and an International Space Station payload. In this talk, we will present the science requirements for the NanoMagSat constellation, alongside the derived system design. A new feature of the NanoMagSat m-NLP is its capability to operate any number of probes in fixed-bias mode, while others are sweeping the bias voltage. This allows for the simultaneous high-resolution determination of the plasma density (2kHz), alongside low-resolution measurements of the electron temperature (a few Hz). Furthermore, the synergy between the in situ plasma density/electron temperature measurements and the magnetic measurements made on board NanoMagSat will be discussed. Finally, initial results from instrument tests within the UiO plasma chamber will be presented.

In the frame of the CubeGrav project, funded by the German Research Foundation, Cube-satellite networks for geodetic Earth observation are investigated on the example of the monitoring of Earth’s gravity field. Satellite gravity missions are an important element of Earth observation from space, because geodynamic processes are frequently related to mass variations and mass transport in the Earth system. As changes in gravity are directly related to mass variability, satellite missions observing the Earth’s time-varying gravity field are a unique tool for observing mass redistribution among the Earth’s system components, including global changes in the water cycle, the cryosphere, and the oceans. The basis for next generation gravity missions (NGGMs) is based on the success of the single satellite missions CHAMP and GOCE as well as the dual-satellite missions GRACE and GRACE-FO launched so far, which are all conventional satellites.
In particular, feasibility as well as economic efficiency play a significant role for future missions, with a focus on increasing spatio-temporal resolution while reducing error effects. The latter include the aliasing of the time-varying gravity fields due to the under-sampling of the geophysical signals and the uncertainties in geophysical background models. The most promising concept for a future gravity field mission from the studies investigated is a dual-pair mission consisting of a polar satellite pair and an inclined (approx. 70°) satellite pair. Since the costs of realizing a double-pair mission with conventional satellites are very high, alternative mission concepts with smaller satellites in the area of New Space are coming into focus. Due to the ongoing miniaturization of satellite buses and potential payload components, the CubeSat platform can be exploited.
The main objective of the CubeGrav project is to derive and investigate for the first time optimized cube-satellite networks for Earth’s gravity field recovery, with special focus on achievable temporal and spatial resolution and reduction of temporal aliasing effects. In order to achieve the overall mission scope the formation of interacting satellites including the inter-satellite ranging measurements, relative navigation of the satellites and networked control of the multi-satellite system are also analyzed in a second step. A prerequisite for the realization of a CubeSat gravity mission is the miniaturization of the key payload, such as the accelerometer, which measure the non-gravitational forces such as the drag of the residual atmosphere, and the instrument for highly accurate determination of the ranges or ranges rates between the satellites.

This contribution presents recent results of the CubeGrav project and a preliminary mission concept and focuses on the scientific added value compared to existing satellite gravity missions. A set of miniaturized gravity-relevant instruments, including accelerometer and the inter-satellite ranging instrument, with realistic error assumptions are identified for a usage in CubeSats, and their capabilities and limits on determining gravity field are investigated in the frame of numerical closed loop simulations. The applicability of the above is further translated into potential preliminary satellite bus compositions and achievable orbital baselines. By this approach we can identify the minimum requirements regarding instrument performances and satellite system design. Additionally, different satellite formations and constellations will be analysed regarding their potential of retrieving the temporal gravity field.

Satellite-based Earth observation data is today the most available it has ever been and still struggles to meet the supply demands from its customers. Meeting end user demand is challenged by conflicting needs such as tasking priority, coverage, quality, spectral band selection, resolution, and product latency. Traditionally priority goes to the highest bidder, leaving emerging applications requiring scientific quality behind or limited to the high-quality government missions. The EarthDaily Satellite Constellation (EDSC) is a customer requirement driven operational enterprise solution for monitoring that works interoperably with government science missions, removes priority tasking by imaging the Earth’s landmass every day, and delivers a flexible scientific-grade product offering designed to seamlessly integrate with machine learning and artificial intelligence algorithms powering geoanalytics applications.

In 2023, EarthDaily Analytics will be launching the 9-satellite EDSC. It will be the world’s first Earth observation system planned, from the ground-up to power machine-learning and artificial intelligence-ready geoanalytics applications on a daily global scale. The processing, calibration and QA engine behind our constellation is the EarthPipelineTM which has been in development for more than eight years and is the world’s first ground segment pipeline as a service. The EarthPipelineTM is our cloud-native processing service that transforms raw downlinked satellite data into high-quality Analysis Ready Data and is designed and tested to handle quality, scale and automation for all sensor types and modalities. This service is based on rigorous satellite and physical modelling, combined with the latest advancements in computer vision and machine learning to automatically produce the highest quality scientific-grade satellite imagery products on the market at scale.

With 20+ spectral bands well aligned with leading science missions, including Sentinel-2 and Landsat-8, and backed by the EarthPipelineTM’s continuous calibration engine, the EarthDaily Mission will be unprecedented monitoring and change detection solution for near real-time situational awareness of the natural environment at scale.

In addition to the global need for environmental stewardship, the market itself demands better monitoring of the environment across all industries due to investor demands and financial risks imposed by climate change and environmental degradation. While open data solutions are more widely available than ever, a persistent need for daily, global scientific quality spectral bands paired with analysis-ready data production remains. Daily scientific data means a chance of cloud-free observation every few days is almost guaranteed and can be used to feed better phenological modelling such as tree carbon accounting and agriculture yield. EDSC includes Short Wave Infrared Bands (SWIR) to dramatically improve landcover differentiation, fire delineation, and improved atmospheric correction and mask generation. Other specialized bands will provide global daily services for scouting the presence of methane anomalies, forest fire detection, impact and risk assessment, water quality evaluation, and carbon cycle monitoring which all serve as a vital input for large-scale climate modelling and mitigation. EDSC’s combination of spectral bands and interoperability with the gold standards of Earth observation (Landsat, Sentinel, and other government science missions) will offer end users an unprecedented combination of daily global coverage, quality and resolution that will deliver impactful solutions to many of the world’s most pressing challenges.

Born from the “Sharing Economy”, OpenConstellation proposes a different way of getting access to satellite imagery. Buying a single satellite, or a small fleet, does not usually provide significant coverage capacity, and the revisit is definitely disappointing. Trying to acquire large, recent coverages in the commercial market proves to be difficult and expensive when quality is a strong requirement. Not to talk about getting real-time imagery in case of emergency, or about the homogeneity problems when using different sources.

Open Cosmos constellation is designed to solve these problems by creating a structure where constellation partners share their spare capacity with the others. Open Cosmos ensures a seamless operation, clear sharing rules and provides all the technical infrastructure, from ground stations to analysis-ready data, in order to make this possible. Becoming a member of the OpenConstellation implies that the investment is immediately multiplied by the sharing effect, and all the constellation management, from downloading to processing and sharing is greatly simplified.

Open Cosmos data platform is the final element in this chain, and provides seamless access to images and derivative products available to partners, downstream providers and final users.

The OpenConstellation is designed to support the services and applications required by its partners, by providing unprecedented access to reliable and affordable satellite imagery and connecting it to a powerful data platform. This allows constellation members to become more efficient and competitive by enabling a full new line of space-data-driven decisions.

The year 2022 will see the first three to five satellites of the constellation launched, and the final schedule for the launch of the full constellation will be fixed.

What is different between the OpenConstellation and other initiatives?

- Cost effective satellite imagery:
Being a member of the OpenConstellation will result in accessing 10x more affordable data.

- The OpenConstellation is born from collaboration:
And thus makes an efficient use of resources by sharing the unused capacity of some satellites to serve other members’ needs.

- The OpenConstellation is built from user needs:
Satellites that become part of the OpenConstellation are coming from actual user demand. No technology push syndrome.

- The OpenConstellation is varied:
Most constellations are based on replicating the same satellite type. The OpenConstellation offers a set of complementary sensors better suited for the demanding new applications. “If you only have a hammer, all problems are nails”.

- Access to the OpenConstellation is provided through a state-of-the-art data platform:
This simplifies the process of finding and managing data, and offers an ecosystem of applications from value-added providers ready to provide standard (e.g. change detection) and bespoke solutions.

- The OpenConstellation evolves quickly with new technological advances:
New sensor technologies are developed every day. The OpenConstellation is enhanced with technological advances much faster than any other due to its open design.

The presentation will describe in depth the constellation design (orbital planes, number of satellites, technical features), the mechanisms designed to effectively implement the capacity sharing, and will demonstrate some early use cases of the data platform run with actual customers.

Observing the Earth and understanding its evolution is fundamentally linked to the ability to collect large amounts of data and extract the needed intelligence to derive and validate the models of such an incredibly dynamic system. While on one side this ability is enabled by the growing number of data sources, on the other side next generation Space platforms, integrating powerful on-board processing capabilities, are bringing transformational opportunities to Earth sciences. As a consequence, on-the-edge computing in Space is becoming a reality thanks to the multiple Teraflops performance of new spacecraft platforms. On top, and in addition to performance, the community is looking for missions with higher spatial and temporal coverage (requiring constellations of satellites) and more rapid design cycles.
LuxSpace has been at the forefront of rapid satellite development already before New Space was a known phrase, developing and building 2 Vesselsats in one year, and the first private-funded Moon mission (4M) in less than half a year.

Today, LuxSpace is developing the innovative Triton-X platform that uses extensive spin-in from automotive and other earth-bound industries to achieve a winning combination of high quality and performance with low recurrent cost and turn-around time. Triton-X will be a modular range of platforms from about 30 to 250 kg, adaptable to a wide range of payloads up to the 100kg class. Triton-X is being developed with the support of ESA, giving LuxSpace access to a huge store of expertise and advise, while keeping the freedom to use New-Space approaches.

Key aspect of Triton-X is the on-board processing power of its integrated avionics unit (IAU) and its robust modular architecture that results in high reliability and robustness on the basis of high-performance low cost high-end-COTS electronics. This architecture can scale easily to different mission sizes and demands, being adaptable in software to the specific needs of the applications. The key elements of this architecture are being developed by LuxSpace and a small core group of partner companies. Due to the low recurring cost of the avionics, Triton-X is especially well suited for small constellations of high-performance satellites.

A number of missions focused on Earth resources monitoring have been studied for potential customers, including atmospheric trace-gas monitoring, maritime surveillance, spectrum monitoring, in-orbit demonstration of a large set of payloads, and others.

Mission Agile Nano-Satellite for Terrestrial Image Services (MANTIS) is a nano-satellite designed to monitor and help understand oil & gas energy supply chains. Oil & gas energy supply chains are highly topical because of their criticality to the development of humankind, but also for the impact they have on nature and the Earth’s climate. They are extremely complex owing to their diversity, distribution and scale. Timely and trustworthy information on these supply chains is valuable to a wide range of user segments such as oil & gas operator and service companies, ESG investors, commodity traders, IGOs & NGOs, and regulators. This poster presentation introduces MANTIS and explains how valuable business insights relating to the oil & gas energy supply chain will be derived from its high-resolution optical imagery.

Satellite remote sensing is a well-established methodology for observing natural and anthropogenic terrestrial and atmospheric processes. The USGS-operated Landsat missions have kept a record of land cover change for 50 years while the Meteosats have observed the atmosphere for a similar period. The Landsats and Meteosats were designed with a clear objective in mind and have since been adapted for solving a wide range of opportunistic goals. Like changing land cover and the atmosphere, understanding something as critical and complex as the energy supply chain warrants a target-specific mentality to mission design. The MANTIS satellite will address a perceived gap in the availability of ultra-economic high-resolution and frequency optical imagery from which oil & gas infrastructure can be detected and classified. These data will be used in concert with a wide range of other Earth Observation (EO) data to derive detailed and timely insights on activity relating to oil & gas production. This topic is addressed in more detail under the ESA ARTES IAP Energy SCOUT project.

Initially, MANTIS will focus on the detection and classification of features and events related to onshore unconventional natural gas production from shale deposits. This form of production, more commonly known as ‘fracking’, is controversial owing to its significant surface footprint impacting on biodiversity, potential impact on the water table, and capacity to release methane, a potent greenhouse gas, into the atmosphere. Unconventional natural gas has however transformed the United States from a net energy importer to exporter (EIA AEO 2020 report), and with significant economically viable shale gas resources known to exist throughout the world, remains a highly topical subject.

[Image: image.png]Historic and forecast energy production and consumption in the United States. EIA Annual Energy Outlook (AEO) 2020.

The unconventional natural gas sector is extremely fast moving with wells being drilled and bought into production in a matter of weeks. Knowing where development and production are occurring and what stage the process is at is critical to understanding how these natural resources contribute to the energy mix and their impact on the environment. The MANTIS mission has been designed to monitor these processes at a spatial and temporal resolution deemed nominal to gain a deeper understanding of activity.

The MANTIS satellite is targeting a 515km, sun-synchronous orbit with a local time of the ascending node (LTAN) of 22:30. Images for the regions of interest will be acquired in the visible (RGB) and near-infrared (NIR) wavelengths. The payload onboard the Mantis mission (iSIM90-12U) offers the same spectral bands specified by ESA Sentinel-2 EO satellites (in the RGB & NIR).

The ground sampling stance (GSD) of the post-processed images, including degradation due to platform and orbital effects, will be 2.5m in RGB and 3.0m in NIR. The images will be characterised by a signal-to-noise ratio of 55 (for a solar elevation angle of 33.8 degrees) and a modulation transfer function (MTF) in the range 17-22%. The mission is being developed to achieve a geolocation accuracy requirement of no more than 100 metres. The aforementioned mission performance has been defined to enable the extraction of valuable information from the MANTIS imagery by means of Terrabotics’ detection and classification workflow.

The Areas of Interest (AOIs) targeted by the mission have been defined considering the regions of highest activity in the unconventionals-based energy supply chain. Short term variations in market demands are also satisfied by autonomous tasking based on inputs from the end user on new Points of Interest (POIs). End users will be able to submit requests for tasking the satellite by submitting these to Open Cosmos’s Mission Operations Centre. These requests will inform the definition of the image acquisition plan.

While imagery data will be available to purchase, the primary use of MANTIS imagery will be providing high-resolution imagery to the Terrabotics Energy SCOUT service. This high-resolution imagery will provide greater information content to Energy SCOUT end users through allowing the identification and classification of small scale events occurring at oil & gas production sites.

SATLANTIS is a European leader in HR and VHR Earth Observation capabilities offering an EO Space Infrastructure built around iSIM (the integrated Standard Imager for Microsatellites) optical-payload concept for small satellites. SATLANTIS offers an End-to-End solution from Upstream satellite development, launch, and operations to Downstream data generation, processing and delivery (e.g., data analytics for methane measurements).

SATLANTIS’ iSIM imager presents three main disruptive capabilities for optical payloads with low mass: enhanced spatial resolution, multispectrality and agility, able to have a relevant impact on several EO value-added applications.

URDANETA is the name of SATLANTIS’ first fully owned satellite, that will be launched in Q2 2022 with SpaceX’s Falcon 9 launcher. This satellite is incorporating an iSIM-90 imager onboard a 16U CubeSat sensor bus providing an innovative solution for several applications on Earth Observation. The main characteristics of URDANETA are: 17.4 kg mass, 4 spectral bands (RGB and NIR), 2.5 m satellite resolution in RGB and NIR, 14.3 km swath, pointing agility: > 1º/sec and 98 Mbps data download rate.

GEI-SAT constellation is the name of a family of four satellites that embarks SATLANTIS’ innovative solution for methane emissions detection and quantification. GEI-SAT pursues hot spot mapping of low emission levels of methane with a low-mass and low-cost very high-resolution multispectral SWIR camera onboard CubeSats & MicroSats.
The constellation consists of a first GEI-SAT Precursor, a 16U CubeSat (17.4 kg, ~150 kg/h detection threshold, 2.2m resolution @VNIR and 13m resolution @SWIR, up to 1700 nm) to be launched in Q3 2023; two more Microsats (92 kg, ~100 kg/h detection threshold, 0.8m resolution @VNIR and 7m resolution @SWIR , up to 1700 nm) to be launched in Q3 2024, and another Microsat expanding spectral capabilities and with similar resolution and better detection threshold (92 kg, ~50 kg/h (TBC) detection threshold, 0.8m resolution @VNIR and 9m resolution @SWIR , up to 2300 nm) to be launched in Q3 2025.

The LEO constellation composed by a CubeSat and three Microsats will employ robust and flight proven platforms compatible with small launchers. The operation lifetime will be of 4 years for the CubeSat, and >5 years for the Microsats. The Ground Segment will include the mission operations & control centre and the data processing and services centre.

The proposed methane detection method is the Multispectral Differential Photometry. It is carried out in collaboration with End users such as ENAGAS by taking images with several filters and, using the different signal values measured at the different wavelengths, obtaining the methane absorption. Before being able to do that, the acquired images have to be corrected for atmospheric effects using radiative transfer models in order to pass from detector units (e-/s) to column concentration units (ppb·m). Again, these concentration units have to be corrected for wind effects, using meteorological models in order to pass from column concentration units to flux units (kg/h).

GEI-SAT constellation will contribute to Improve annual reporting on methane emission with higher frequency measurements & prepare for global certification on CH4 emissions reduction in future legislation world-wide. The high spatial resolution that GEI-SAT provides in its SWIR channel, together with the geo-localization provided by its very high resolution for VNIR channel images, will allow an unprecedented ability (4 to 16 times better than other satellites) to distinguish the exact location of the methane leak or uncontrolled emission at a global scale. In order to achieve this object, the constellation will be operated in coordination with Sentinel-5P to perform an operational Tipping and Cueing capability for quantification of point source of CH4.

By enhancing EO capabilities from Space and pursuing climate Mitigation applications deriving from satellite-based observations, both satellite missions are based on innovative EO data processing from space and share the goals of a more sustainable life on Earth (showing the end-to-end service provider including space+ground segments), using SATLANTIS’ own innovative imagery pinpointing to specific application domains such as GHG emissions (GEI-SAT) and lands/water quality (URDANETA).

The BlackSky constellation, owned and operated by the US company BlackSky Inc. (NYSE: BSKY), whose data is distributed by Telespazio/e-GEOS in Europe and world-wide thanks to an agreement signed on 2018, is a new EO very high resolution optical (VHRO) mission designed with the goal of providing the highest daily revisit on the market at about 1mt resolution.
Currently the system is composed by eight satellites, launched since 2019; the two newest satellites, launched on November 18th, 2021, reached orbit and delivered first images within 14 hours of launch. The constellation is growing fast: additional two/four satellites are planned for launch within end of 2021, to be followed by other satellites of the same class (less than 60 kg mass) by mid 2022. The next achievement of a baseline of at least 16 operational satellites will allow a revisit of more than 8 acquisitions every day.
The orbital configurations of the constellation (including polar and inclined orbits) allows to achieve multiple acquisitions every day, during all daylight hours, with on-demand satellite tasking and fast access to the constellation available at multiple priority levels, granting a unique “First-to-know” advantage.
The imaging performance is based on a framing camera with color filter array that, in its latest version (thanks also to the position to a lower orbit), provides a sub-metric resolution over a n about 25 Km2 area.
The BlackSky image quality is monitored and calibrated continuously, as soon as new satellites become available, using specific calibration targets. e-GEOS will present some of its internal analysis on the images, as well as examples of operation tests that have already demonstrated a very short tasking lead-time and fast delivery timelines.

Gravity waves are important for atmospheric dynamics and play a major role in the mesosphere and lower thermosphere (MLT). Thus, global observations of gravity waves in this region are of particular interest. To resolve the upward propagation, a limb sounding observing system with high vertical resolution is developed to retrieve vertical temperature profiles in the MLT region. The derived temperature fields can be subsequently used to determine wave parameters.

The measurement method is a variant of Fourier transform spectroscopy: a spatial heterodyne interferometer is used to resolve rotational structures of the O$_2$ atmospheric A-band airglow emission in the near-infrared. It is visible during day- and night-time, allowing a continuous observation. The image is taken by a 2d detector plane containing of hundreds of pixels on one axis. The horizontal axis contains the interference pattern. The vertical axis corresponds to different tangent altitudes. Thus, it allows to resolve a vertical profile with one image. The method exploits the relative intensities of the emission lines to retrieve temperature. Thus, no absolute radiometric calibration is needed, which facilitates the calibration of the instrument. Silicon-based detectors, like CCD or CMOS, can be used. These can operate in ambient conditions and do not require active cooling devices. This allows to deploy this instrument on a nano or micro satellite platform such as CubeSats.

After a successful in-orbit demonstration of the measurement technology in 2018, this instrument will be developed next within the International Satellite Program in Research and Education (INSPIRE). The European Commission has preselected the instrument for an in-orbit validation to demonstrate innovative space technologies within its H2020 program.

Following the success of the PHI-Sat mission, in 2020, the European Space Agency (ESA) announced the opportunity to present CubeSat-based ideas for the PHI-Sat-2 mission to promote innovative technologies such as Artificial Intelligence (AI) capabilities onboard Earth Observation (EO) missions.
The PHI-Sat-2 mission-idea, submitted jointly by Open Cosmos and CGI, leverages the latest research and developments in the European ecosystem: a game-changing EO CubeSat platform capable of running AI Apps that can be developed, uploaded and deployed and orchestrated on the spacecraft and updated during flight operations. This approach allows continuous improvement of the AI model parameters using the very same images acquired by the satellite.
The Development is divided into two sequential phases: the Mission Concept Phase, which is almost completed, which shall demonstrate the readiness of the mission by demonstrating the innovative EO application through a breadboard base validation test and a Mission Development Phase, which shall be dedicated to the design and development of the space and ground segments, launch, in-orbit operations, data exploitation, and distribution.
The PHI-Sat-2 Mission, lead by OpenCosmos, will be used to demonstrate the AI enabling capability for new useful innovative EO techniques of relevance to EO user communities. The overall objective is to address innovative mission concepts, fostering novel architectures to meet user-driven science and applications by means of on-board processing. The latter will be based on state-of-the-art AI techniques and on-board AI-accelerator processors
The mission will take advantage of the latest research for mission operations of CubeSats and use the NanoSat MO Framework, that allows software to be deployed in space as simple Apps, in a similar fashion to Android apps, previously demonstrated in ESA’s OPS-SAT mission, and supports the orchestration of on-board Apps.
Φ-sat-2 will a set of default AI Apps, which will cover different ML approaches and methodologies such as supervised (image segmentation, object detection) and unsupervised learning (with auto encoders and generative networks) and presented below.
Since the Φ-sat-2 mission relies on an optical sensor, the availability of a Cloud Detection App –Develop by KP-Labs- which will generate cloud mask and identify cloud free areas is a baseline. But this information can be exploited by the other Apps and this is not only relevant for onboard resources optimization, but also will demonstrate the AI Apps pipeline onboard;
Autonomous Vessel Awareness App –developed by CEiiA- - will detect and classify vessels. Together with the demonstration of the possibility to perform scouting with a sensor with wider swath width, this App will demonstrate how information generated in space can be exploited for mission operation e.g. in a satellite constellation to identify areas for the next acquisitions.
The Sat2Map App – developed by CGI- transforms a satellite image to a street map in emergency field using Artificial Intelligence. The software takes advantage of the Cycle-Consistent Adversarial Networks (CycleGAN) technique to do the transformation from the satellite image to the street map. This App will enable the satellite to provide to rescue teams on ground in case of emergency (Earthquake, flood etc.) real time of still available and accessible street.
High Compression App developed by Geo-K- will exploit deep auto encoders to push AI image compression on-board and on ground reconstruction. The performances of the App will not be measured only in terms of standard compression rate VS image similarity but also in terms of how the reconstructed image will be exploitable by other Apps e.g. for objects recognition pushing the limit of image compression in space based on AI and reconstruction on the ground.
On top of this, the mission will be open to applications developed by third parties and this augment the disruptiveness of a new mission concept where the Satellite can be seen available to a community already in space for research and development as a commodity. Those third party APPs can then be uploaded and started/stopped on demand. This concept is extremely powerful enabling future AI software to be developed and easily deployed in the spacecraft. This will represent an enabler for in-flight on-mission continuous learning of the AI networks.
The presentation aims to describe the PHI-Sat-2 mission objectives and how the different AI applications, orchestrated by the NanoSat MO Framework, will demonstrate the disruptive advantages that the onboard AI brings to the mission.