The Earth Cloud, Aerosol and Radiation Explorer (EarthCARE) mission, is the sixth Earth Explorer Mission of the European Space Agency ESA in cooperation with the Japan Aerospace Exploration Agency (JAXA). With its payload, consisting of 2 active and 2 passive instruments, it will provide data to improve the understanding of the processes involving clouds, aerosols and radiation in the Earth’s atmosphere, addressing uncertainties in global models for climate predictions but also strong weather events in numerical weather prediction.

Flying in a sun-synchronous orbit at 393 km altitude with a 14:00 descending node, the mission produces data from the 4 co-aligned instruments, an Atmospheric UV LIDar (ATLID), a Multi-Spectral Imager (MSI), a Broad-Band Radiometer (BBR) and JAXA’s Cloud Profiling Radar (CPR). This payload data is processed individually and in a synergetic manner by the ground segment into a large number of EarthCARE data products which include among others vertical profiles of aerosols, liquid water and ice, observations of cloud distribution and vertical motion within clouds.

After years of development, manufacturing, integration and testing, and overcoming technological and technical challenges, the EarthCARE satellite has been integrated and functionally tested at the facilities of the prime contractor, Airbus Defence and Space GmbH in Friedrichshafen (Figure 1). It is currently being prepared for the environmental test campaign. All teams, covering the satellite, launcher, operations, ground segment, mission data processing and products generation, validation and scientific exploitation are currently in full preparation for the EarthCARE launch in 2023 and subsequent commissioning and exploitation phases.

This paper will present a general overview of the EarthCARE mission, its main objectives, requirements and design. It will give a general view of the past and present development, assembly and testing activities for the EarthCARE space and ground segments, validation and science preparations and will address the main challenges encountered and associated lessons learned. The current status and most importantly, the plans for the remaining activities towards launch, commissioning and exploitation will be described in detail.

EarthCARE CPR has been developed by Japan Aerospace Exploration Agency (JAXA) and National Institute of Information and Communications Technology (NICT) for the Earth Clouds, Aerosols and Radiation Explorer (EarthCARE) mission. This is the world’s first cloud radar that measures vertical profiles of the cloud and its vertical motion.
The CPR is a 94GHz pulse radar and measures altitude and Doppler velocity from received echo. Its minimum sensitivity is greater than -35dBZ and the Doppler velocity measurement accuracy is less than 1.3m/s. To achieve those performances, the CPR equips a big reflector with diameter of 2.5m. The observation altitude can be changed with Low (16 km), Middle（18 km), High (20 km).
CPR was transported to Europe in March 2021 and was handed over to ESA/Airbus in April 2022. Then CPR integration to EarthCARE satellite and test was completed by ESA/Airbus in June 2021.
This presentation introduces CPR overview and the latest status of preparation towards launch.

The EarthCARE payload consists of two active and two passive instruments that will provide collocated images, enabling synergistic use of the data that will be used to derive information about atmospheric processes. ESA has developed three of the instruments, an ATmospheric LIDar (ATLID), a Multi-Spectral Imager (MSI) and a Broad-Band Radiometer (BBR). JAXA has developed a Cloud Profiling Radar (CPR).

The objective of the BBR is to derive instantaneous, broadband, top of atmosphere fluxes, with an accuracy better than 10 Wm-2. The instrument development was led by TAS-UK, with an Optics Unit from RAL (UK). 10 x 10 km scenes fore, nadir and aft on the satellite ground track are obtained, with measurements made in shortwave and longwave channels, using three telescopes, each with a linear, microbolometer array, operating with push-broom motion. Thereby, matched short and long wave scenes are provided from the three telescope views, that are spatially coincident, from different observation angles, and with small separations in observation time. Whilst a 10 km scene size defines the performance requirements, BBR field of view is configurable and can therefore also be processed, for instance, in a 21 km long and 5 km wide configuration (21 km due to CPR cycles of 7 km length). On-board calibration is performed using views to blackbodies and a solar illuminated diffuser.

The objective of MSI is to provide contextual information on the horizontal structures of clouds, more specifically cloud type and cloud optical and microphysical properties over sea and land surfaces. Further, it has a goal to provide information concerning aerosol over sea surfaces, to supplement the ATLID aerosol measurements in an across-track dimension. The instrument development was led by SSTL (UK), which also built the three channel Thermal Infrared (TIR) camera, whilst a four channel Visible, Near infrared, Short-wave infrared (VNS) camera has been built by TNO (NL). The instrument collects data over a 150 km swath that is pointed slightly across track in order to reduce sun glint. The VNS camera operates in pushbroom, to collect data over four focal planes employing linear photodiode arrays. Calibration is via closed shutter dark images and a view to a solar illuminated diffuser. The TIR camera is equipped with a 2-D microbolometer array and uses Time Delay Integration in order to increase the signal to noise ratio. Calibration is via a blackbody and a view to cold space.

ATLID objective is the provision of vertical profiles of optically thin cloud and aerosol layers, characterising aerosol optical properties and measuring the altitude of cloud boundaries while detecting small ice particles and water droplets. It will complement cloud observations from the CPR. Airbus Toulouse (F) has led the development of ATLID, with the laser transmitter assembly integrated by Leonardo (I). The instrument is a bistatic, high spectral resolution lidar that emits short laser pulses in the UV, at a repetition rate of 51 Hz. The 620 mm aperture receiving telescope filters the return signal through the optics of the instrument focal plane assembly, separating the signals backscattered by the atmosphere to provide measurements of the aerosol (Mie) and molecular (Rayleigh) components. The Mie co and cross-polarised components are also distinguished. After initial calibration on orbit ensures best alignment between the transmit and receive channels and best receiver focus, a coarse spectral calibration is performed. A fine spectral calibration is then regularly undertaken, as well as calibrations to monitor the detectors, spectral and polarisation cross talk and lidar constants.

This paper will provide an overview of the design and function of the instruments with results from their on-ground calibration and their performance predictions.

The interactions between clouds, aerosols and solar and terrestrial radiation play key roles in the Earth’s Climate. It has been recognized that, despite a long history of satellite observations, further high-quality novel observations are needed for atmospheric model evaluation and process studies. In particular, the importance of true height-resolved global observations of cloud and aerosol properties has been recognized as being essential to making progress. EarthCARE is an upcoming ESA/JAXA mission to fly in 2023 which will focus on making these observation.
The four EarthCARE instruments provide synergistic observations of cloud and aerosol profiles, precipitation and broad-band solar and thermal fluxes are:

• A 94 GHz, Doppler Cloud Radar (CPR) supplied by JAXA which will be the first 94 GHz radar in space with Doppler capability, to measure (thick) cloud profiles, vertical ice particle velocities and precipitation.
• An advanced 355 nm High-Spectral Resolution Lidar (ATLID) including a total depolarization channel
• A Multispectral imager for narrow-band TOA radiances (MSI) to provide across-track scene context information and additional cloud and aerosol information.
• A 3-view Broad-Band Long- and Short-Wave Radiometer for TOA radiance (BBR) to measure the outgoing emitted, respectively, reflected, broad-band solar and thermal radiation at the top of the atmosphere.

The ESA scientific retrieval processors are fully exploiting the synergy of these observations. EarthCARE will provide twenty-five science (Level 2) products generated by seventeen separate processors. These products include nadir profiles of cloud, aerosol and precipitation properties along with constructed three-dimensional cloud-aerosol-precipitation domains and associated with calculated radiative properties, such as heating rates The final L2 processor compares the forward modelled top-of-atmosphere broad-band radiances and fluxes based on the constructed 3D atmospheric scenes with those measured by the BBR in order to assess and improve the quantitative understanding of the role of clouds and aerosols on radiation. In the autumn of 2021 the seventeen individual processors were chained together for the first time to simulate the full processing chain from input L1 signals to the final L2 retrievals.
The presentation will provide an overview of the scientific data products, processors and preparations for in-orbit validation.

Earth Clouds, Aerosols and Radiation Explorer (EarthCARE) mission is designed to produce the maximum synergetic collaboration of European and Japanese science teams. The EarthCARE products will be developed and distributed from both JAXA and ESA. Continuous exchanges of information have been conducted between Japan and Europe through the Joint Algorithm Development Endeavor (JADE). CPR, ATLID and MSI Level-2 provide cloud mask, cloud phase and cloud microphysics (such as cloud effective radius, liquid water content, optical depth, etc) for the respective sensor products, together with the synergy products by using the combination of the sensors. Further, the CPR provides the Doppler velocity measurement (which gives the vertical information of the in-cloud velocity), and precipitation products. ATLID Level-2 includes aerosol flagging, aerosol component type (such as dust, black carbon, sea salt and water soluble), as well as the aerosol optical properties including aerosol extinction. The cloud and aerosol products will be used to derive the radiative flux at shortwave and longwave, whose consistency with the BBR will be checked to produce the final radiation product by 4-sensors.
Validation activities are necessary to distribute the scientific products whose quality and reliability are assured. The JAXA is planning the validation activities by utilization of the existing observation network, campaign observation, and cross comparison with other satellite data.
Furthermore, a wide range of application research activities will be planned to achieve the mission objectives. EarthCARE observation data will contribute to understanding cloud, aerosol, and radiation processes, evaluations and improvements of climate models and numerical weather prediction (NWP) models, and atmospheric quality monitoring. The Intergovernmental Panel on Climate Change (IPCC) report published in August 2021, “Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the IPCC”, summarizes that the cloud feedback remains the largest contribution to overall uncertainty, and contributions to mitigate the uncertainty can be expected by new insights by the EarthCARE observations.
This presentation will introduce JAXA Level 2, Validation and Applications Preparation.

The EarthCARE mission, implemented in cooperation with JAXA, will be the largest and most complex ESA Earth Explorer mission built to date, and its products will contribute fundamentally to the understanding of the climate system. The combination of two active (lidar and radar) and two passive instruments (imager and radiometer) instruments will provide synergistic observations of cloud and aerosol profiles, precipitation and broad-band solar and thermal fluxes.

ESA and JAXA defined and are coordinating a joint EarthCARE Scientific Validation Implementation Plan. This presentation will then focus on the ESA-coordinated Validation activities, in particular on validation of the Level 1 products of the ESA instruments (the atmospheric lidar instrument (i.e. ATLID), the broad band radiometer (i.e. BBR), and the multi-spectral imager (i.e. MSI)) and on the ESA-developed Level 2 products. These ESA Validation activities have been the outcome of an ESA announcement of opportunity that was issued in 2017 and for which more than 30 proposals had been received. A broad peer review of this program took place in 2018 during the 1st ESA Validation Workshop in Bonn (held in concomitance with the 7th EarthCARE Science Workshop), to assess the scope of the proposed activities. A second workshop was held online in March May 2021 to review the validation approaches and methods. Here, also the broader context was addressed, with EarthCARE products contributing to space-borne Earth observation data record together with those with from earlier and later missions /instruments such as Aeolus, Calipso, Cloudsat, CERES, GPM, Aeolus Follow-On, and ACCP/AOS. Many of the workshop recommendations concerned common practice consolidation for aerosol and cloud profile validation, which will be addressed in a dedicated poster.

The EarthCARE product validation will begin during the 6-month commissioning phase and will continue during the entire exploitation phase of at least 2.5 years

In preparation of this exploitation phase, ESA intends to foster EarthCARE application development via its Research Opportunity announcements under its broader Atmospheric Science Cluster scheme. The presentation will give very brief guidance on this mechanism, in particular the schedule of open and upcoming calls.

A further preparation activity for exploitation of EarthCARE data that is already well underway is aimed at achieving readiness for EarthCARE data assimilation.: The high-resolution, profiling observations of clouds from EarthCARE will contain a wealth of information on the current atmospheric state and therefore have the potential to improve the initialisation of weather forecasts. To fully exploit this, ESA and the European Centre for Medium-Range Forecasts (ECMWF) have been working closely to ensure that EarthCARE’s observations can be assimilated at a global numerical weather prediction centre as soon as possible after launch. Observing system experiments where CloudSat radar reflectivity and CALIPSO lidar backscatter are assimilated in the ECMWF integrated forecast system (IFS) have revealed the power of EarthCARE’s novel observations to have a direct benefit on forecasts of temperature, humidity and winds. Including EarthCARE data within the IFS will also allow the data to be monitored against model forecasts and thus provides an invaluable validation tool for the rapid detection and diagnosis of observation issues.

Antarctic ice shelves border c. 75% of the Antarctic coastline, and often act to buffer grounded ice contributions to sea level rise. The mass balance and stability of ice shelves is partly determined by surface melting, which can lead to the formation of extensive surface meltwater systems that may result in hydrofracturing and eventual ice shelf disintegration. It is crucial, therefore, that we better understand the extent, onset, and duration of surface meltwater systems across Antarctic ice shelves. Surface meltwater systems comprise slush (saturated firn), ponded meltwater (lakes) and streams. Meltwater ponding may increase an ice shelf’s vulnerability to hydrofracture whereas runoff in streams may reduce it. Where meltwater re-freezes within the firn pack over successive seasons, it can drive firn densification, encouraging further surface ponding and therefore increased vulnerability to hydrofracture.

Here we use two random forest classifiers, trained separately for application to Sentinel-2 and Landsat 8 optical images, to map both slush and ponded water across Antarctic ice shelves from 2013 to 2021. Early results across the Antarctic Peninsula (AP) show marked spatial and inter-annual variability, with peaks in surface meltwater extent observed predominantly during the 2017/2018 and 2019/2020 melt seasons on the western AP ice shelves, but during the 2016/2017 melt season on the eastern AP ice shelves. Intra-annually, the extent of ponded water typically peaks in January-March, and the maximum extent of slush typically co-insides with or precedes this. Across most ice shelves, surface meltwater is observed towards the grounding line, close to areas of exposed bedrock and blue ice. However, on George VI Ice Shelf, surface meltwater extent is more extensive, and is observed across much of the northern portion of the ice shelf in the 2019/2020 melt season.

Icebergs impact the physical and biological properties of the ocean along their drift trajectory by releasing cold fresh meltwater and nutrients. This facilitates sea ice formation, fosters biological production and influences the local ocean circulation. The intensity of the impact depends on the amount of meltwater. A68 was the sixth largest iceberg ever recorded in satellite observations, and hence had a significant potential to impact its environment. It calved from the Larsen-C Ice Shelf in July 2017, drifted through the Weddell and Scotia Sea and approached South Georgia at the end of 2020. Finally, it disintegrated near South Georgia in early 2021. Although this is a common trajectory for Antarctic icebergs, the sheer size of A68A elevates its potential to impact ecosystems around South Georgia through release of fresh water and nutrients, through blockage and through collision with the benthic habitat.

In this study we combine satellite imagery data from Sentinel 1, Sentinel 3 and MODIS and satellite altimetry from CryoSat-2 and ICESat-2 to chart changes in the A68A iceberg’s area, freeboard, thickness, volume and mass over its lifetime to assess its disintegration and melt rate in different environments. We find that A68A thinned from 235 ± 9 to 168 ± 10 m, on average, and lost 802 ± 35 Gt of ice in 3.5 years. While the majority of this loss is due to fragmentation into smaller icebergs, which do not melt instantly, 254 ± 17 billion tons are released through melting at the iceberg’s base - a lower bound estimate for the fresh water input into the ocean. Basal melting peaked at 7.2 ± 2.3 m/month in the Northern Scotia Sea. In the vicinity of South Georgia we estimate that 152 ± 61 Gt of freshwater were released over 96 days, potentially altering the local ocean properties, plankton occurrence and conditions for predators. The iceberg may also have scoured the sea floor briefly. Our detailed maps of the A68A iceberg thickness change will be useful to investigate the impact on the Larsen-C Ice Shelf, and for more detailed studies on the effects of meltwater and nutrients released off South Georgia. Our results could also help to model the disintegration of other large tabular icebergs that take a similar path and to include their impact in ocean models.

The quantification of the sea ice mass balance as the marine part of the cryosphere depends on sea ice thickness satellite observations for the entire ice-covered oceans. The challenges to this task are numerous. Sea ice itself is a highly dynamic medium with a significant variability at meter scale and a strong seasonal cycle which significantly impacts it remote sensing signature. Satellite sensors must therefore provide precise observations at high spatial resolution to observe the full spread of the sea ice thickness distribution and its governing processes. Average thickness values for larger areas are sufficient for mass balance estimates, but available methods such as satellite altimetry and passive microwave remote sensing rely on indirect methods and auxiliary information. Even more challenging is the estimation of sea ice thickness during the presence of surface melt. In addition, suitable satellite sensors in orbits that enabling sea ice thickness retrieval in the inner Arctic Ocean have been in service only until recently in comparison to satellites capable of observing sea ice area. Thus, the assessment of the sea ice mass balance for longer time series is often based on reanalysis models and not Earth Observation data since satellite thickness data records observing the full sea ice cover are short and more than often do not provide information during the minimum of the annual sea ice cycle.

But when available the parameter for the mass budget is traditionally sea ice volume and mass. We will therefore present an available sea ice volume data record that is derived by data fusion of CryoSat-2 radar altimeter and SMOS L-Band passive microwave-based sea ice thickness information. Both methods have a complementary sensitivity to different thickness classes and optimal interpolation is employed for gap-less sea ice thickness information in the northern hemisphere since November 2010. The data record is generated for the ESA funded SMOS & CryoSat-2 Sea Ice Data Product Processing and Dissemination Service (CS2SMOS-PDS).

We discuss the characteristics of the data set and provide an overview of intended evolutions of the data set, specifically improvements to the spatial resolutions, a potential extension to the southern hemisphere and the addition of other available satellite sensors to the optimal interpolation. Within the context of the mass balance of the cryosphere we will share our thoughts on the significance of the CryoSat-2/SMOS based sea ice volume time series for climate applications in the context of its comparable short temporal and how this information can be presented more consistently to other components of the cryosphere.

Ice shelf calving is an important mass loss component of the Antarctic Ice Sheet (AIS). The discharge of the AIS is sensitive to changes in ice shelf extent and thickness due to ice shelf buttressing effects. Up to now, it has been challenging to continuously monitor calving front change as the manual delineation of ice shelf fronts is very time-consuming and suitable satellite data was missing. Since the launch of Sentinel-1, plenty of imagery over the Antarctic coastline is freely available. To overcome the tedious manual work, we present a novel ice shelf monitoring service called “IceLines” which enables the near-real-time tracking of front positions based on Sentinel-1 SAR data. IceLines is based on a Deep Learning (DL) algorithm to automatically extract the border between ocean and ice sheet from high volumes of Sentinel-1 data. Further processing creates a line shapefile for each front position based on suitable Sentinel-1 acquisitions to create a time series of calving front movement. A truncation algorithm secures that erroneously extracted fronts due to surface melt, wind roughening sea or dry snow are excluded from the time series. For validation, manually delineated fronts for randomly selected dates (outside the training data time period) are created. The accuracy is measured as the mean and median distance between DL-extracted and manually delineated fronts. Finally, the continuously extracted calving fronts are made available to the public via the DLR GeoService (geoservice.dlr.de). The webservice will provide researchers with up-to-date front positions on major Antarctic ice shelf extents. This will allow to easily include ice shelf front fluctuations in future analysis in order to provide more accurate estimates on the mass balance of the AIS.

Ice shelves play a crucial role in controlling rates of ice discharge across Antarctica’s grounding lines. Mass loss from ice shelves, predominately due to basal melting and calving, can reduce the buttressing force provided by ice shelves, leading to increased grounded ice discharge. Despite the importance of ice shelves, existing estimates of calving and freshwater fluxes from ice shelves have utilised disparate datasets valid for inconsistent time periods or have relied on invalid assumptions, resulting in a limited account of the health of many ice shelves and little indication of processes driving ice shelf mass imbalance. Here, we quantify these fluxes at annual temporal resolution during 2010 to 2019. On average during the study period, a calving flux of 1283±109 Gt yr-1 balances a melt flux of 1247±149 Gt yr-1. Inter-annual variations in the fluxes of both basal meltwater and calving mean that the melt contribution to ice shelf mass loss varies between 35% and 62%, with the lowest contributions in years with large calving events. These large (>100 Gt) calving events are rare (8 events during 2010-2019), yet account for 35% of the total ice shelf calving flux, highlighting the importance of large calving events for ice shelf mass balance. Eighty percent of ice shelves, including many in East Antarctica, are melting at or faster than their balance rates, indicating that ocean-driven erosion of ice shelf grounding lines is widespread around Antarctica. Furthermore, we find a significant and strong positive correlation (R=0.68) between basal melt flux and grounding line discharge, implying that ocean-driven melt may pace grounded ice loss from Antarctica.

Sea ice is a key component of the earth’s climate system as it modulates the energy exchanges and associated feedback processes at the air-sea interface in polar regions. These exchanges strongly depend on openings in the sea-ice cover, which are associated with fine-scale sea-ice deformations. Viscous-plastic sea-ice rheologies, used in most numerical models, struggle to represent these fine-scale sea-ice dynamics without going to very costly horizontal resolutions (~1km). One solution is to use a rheological framework based on brittle mechanics associated with damage propagation to simulate such deformations. This approach enables to reproduce the characteristics of observed sea-ice deformations with little dependency on the mesh resolution. Here we present results from the first ocean--sea-ice coupled model that uses such a rheological framework. The sea-ice component is given by the neXt generation Sea-Ice Model (neXtSIM), while the ocean component is adopted from the Nucleus for European Modelling of the Ocean (NEMO).

Results for the period 2000-2018 are evaluated using remote sensing observation datasets for each metric of interest: the OSI-SAF drift and concentration datasets for ice drift and extent, the CS2SMOS ice thickness dataset for ice volume and the RGPS dataset for sea-ice deformations. We also evaluate the winter ice mass balance of the model using a recent dataset of sea-ice volume changes estimated using version 2.0 of the ESA CCI sea-ice thickness dataset combined to the Centre ERS d’Archivage et de Traitement (CERSAT) sea-ice motion dataset. We find that sea-ice dynamics are well represented in the model, showing a remarkable match with satellite observations from large scales (sea-ice drift) to small scales (sea-ice deformation). Other sea-ice properties relevant for climate, i.e., volume and extent, also show a good match with satellite observations. We assess the relative contribution of dynamical vs. thermodynamic processes to the sea-ice mass balance in the Arctic Basin and find a good agreement with ice volume changes estimated from the ESA CCI sea-ice thickness dataset in the winter, especially for the dynamical contribution.

Using the unique capability of the model to reproduce sea-ice deformations, we estimate the contribution of leads and polynyas to winter ice production. We find that ice formation in leads and polynyas adds up 25% to 40% of the total ice growth in pack ice in winter, showing a significant increase over the 18 years covered by the model simulation. This coupled framework opens new opportunities to understand and quantify the interplay between small-scale sea-ice dynamics and ocean properties that cannot be inferred from satellite observations.

Title: Unlocking the Power of HAPS for Earth Observation

The HAPS Alliance is an industry association of High-Altitude Platform Station (HAPS) industry leaders that include telecommunications, technology, aviation, and aerospace companies, as well as public and educational institutions. United by a vision to address diverse social issues and create new value through the utilization of high-altitude vehicles in the stratosphere, the Alliance is working to accelerate the development and commercial adoption of HAPS technology by promoting and building industry-wide standards, interoperability guidelines and regulatory policies in both the telecommunication and aviation industries.

In addition to connectivity applications to help bridge the digital divide, HAPS is being used for Earth observation. Fitted with Earth observations payloads, HAPS-enabled sophisticated, uncrewed, high altitude long-endurance vehicles, free balloons, and airships are demonstrating in-flight tests how HAPS can deliver high-quality imagery and video continuously the stratosphere. In the future, these HAPS-enabled flights will enable users to deliver real-time situational awareness, gather data around disasters, support effective rescue responses, predict highly accurate weather reports, and more. This presentation will introduce the HAPS Alliance vision for HAPS operations at scale, provide insight into our approach to Earth observation, and share examples of member progress to date. Some actual use cases and tests to be covered will include the following.

HAPS Alliance Member Sceye’s wind-driven lighter-than-air HAPS platform to provide connectivity and Earth observation. Sceye partnered with the EPA to study pollution sources and their impacts on climate and air quality. In a recent connectivity test, Sceye demonstrated the ability to connect with LTE devices up to 120km horizontal distance, demonstrating the large potential that HAPS can offer over terrestrial infrastructures.
HAPS Alliance Member Raven Aerostar demonstrated this year the utility of their free-flying Aerostar balloon in the stratosphere to support firefighting efforts over the Western United States. Raven Aerostar’s Thunderhead Balloon Systems® offers a high state of technical readiness, well-developed balloon manufacturing capability and experienced flight operations crews.
HAPS Alliance Member HAPSMobile, a SoftBank majority-owned joint venture with AeroVironment, achieved a 5 hour and 38-minute flight in the stratosphere with its fixed-wing solar-powered Sunglider aircraft. Using smartphones connected to the Internet through Sunglider’s LTE payload in the stratosphere, and successfully made a video call to HAPSMobile members based in Japan.
HAPS Alliance Member Deutsche Telekom, a SoftBank majority-owned joint venture with AeroVironment, achieved a 5 hour and 38-minute flight in the stratosphere with its fixed-wing solar-powered Sunglider aircraft. Using smartphones connected to the Internet through Sunglider’s LTE payload in the stratosphere, members from Loon and AeroVironment in the US successfully made a video call to HAPSMobile members based in Japan.
HAPS Alliance Member Airbus completed its 2021 Zephyr flight campaign in Arizona to demonstrate how its “Carbon Neutral” Zephyr can remain in the stratosphere for days and months at a time, showing precision and re-tasking flexibility in the stratosphere.

This presentation will also look at other HAPS applications besides Earth observation, developments in the HAPS industry, and the activities of HAPS Alliance member companies.

High Altitude Pseudo-Satellites (HAPS) are unmanned vehicles continuously flying in the Stratosphere or near-Space for months at a time. When equipped with Earth Observation (EO) or Telecommunication payloads, they provide new service possibilities that complement existing satellite and airborne solutions. Like satellites in the 60s and 70s, HAPS technology development requires years of R&D, flight trials and continual investment. However, the HAPS landscape is now changing quickly.

Recent industry achievements demonstrate that HAPS EO services are already operational, reliable and scalable. The last pending frontier for the aerospace industry is being effectively crossed. The high participation in the recent HAPS Alliance Summit demonstrates that expectations on HAPS solutions are growing, not only for the new service provisions but also because HAPS are inherently a green technology, fully aligned with current international commitments towards a sustainable future.

This paper presents Airbus’ successful experience with Zephyr, the company’s persistent fixed-wing heavier-than-air HAPS, applied to Earth Observation. The growing experience in Zephyr EO campaign preparation and execution highlights the operational similarities and differences between HAPS, satellites and airborne platforms. Airbus benefits from its past and present heritage in both fields: air and space. Aircraft tasking, sensor operation and data transfer, processing, analytics and dissemination, are key aspects to provide consistent services to end users.

Building on the two successful Zephyr stratospheric flights in 2021, the paper will share how some of the demonstrated capabilities and advancements will be instrumental to the benefit of environmental applications and to service civil society, among others.

For the wildfires domain, for instance, data from EO satellites is successfully used to assess fire risk, to calculate burnt surfaces and even as an independent, homogeneous means to register fires worldwide. However, current satellites often do not detect fires when they are still small and controllable and do not provide frequent flame progress updates as required by firefighters. A constellation of HAPS adequately deployed over high-risk areas will enable early fire detection and 24/7 persistent monitoring of active fires, complementing or even replacing the current manned daylight-only airborne surveillance.

Similarly, air quality monitoring in and around urban and industrial areas will benefit from the regional coverage of HAPS, their low revisit time and the duration of their missions. The rapid re-tasking and mobilisation of a flying HAPS will contribute to disaster relief activities offering reactivity and significantly better image resolution (GSD) than current satellites. Earth Observation combined with cell phone or radio relay is also a highly valued capability for emergency services acting in remote or not-well-covered areas. Maritime HAPS applications like illegal fishing control or oil spill detection, combining imaging with AIS/VDES processing, will complement what is currently achieved by satellites and aircrafts.

Data captured from HAPS emerge as a new, valuable source in the integrated, multi-layered nature of geospatial data servicing the Earth Observation and Scientific communities, and will definitely contribute to the achievement of ESA’s and EU’s objectives in regards to our “Living Planet” and the UN’s Sustainable Development Goals (SDGs).

Finally, the paper will provide an insight on the path followed by ESA and the HAPS industry towards a first long-duration operational HAPS demonstration in the EU territory.

Poor air quality (AQ) is a health issue in both develop and developing countries, particularly in urban areas. Cities currently encompass most of the population and are foci of air pollution from industries, household heating/cooling, and traffic. Exposure to noxious gases or small particles is statistically and medically proven to cause lung diseases and premature deaths. Cities also account for more than 70% of the anthropogenic CO2 emissions. The Intergovernmental Panel on Climate Change (IPCC) concluded that human-produced greenhouse gases (GHG) such as carbon dioxide, methane and nitrous oxide are inexorably driving the observed increase in Earth's temperatures observed over the past 50 years. The Paris Agreement made the verification and improvement of local GHG emission inventories imperative.

Monitoring emissions and air pollution concentrations over urban areas requires data granularity at the local level, better than that provided by current and planned satellite missions and ground networks: horizontal and vertical resolutions do not always fit the observational requirements to be used in combination with urban and local air quality models. Urban air quality stations have a sparse coverage and local observations suffer from a limited representativeness. Moreover, the stations network density decreases towards suburbs and adjacent rural sites, hampering a citywide instantaneous view of air quality and the attribution of the pollution to its sources. On the other side while satellite observations are suitable to provide AQ information on a global and regional scale, they have limited capability to provide information at urban and local scale. In response to these challenges High Altitude Pseudo Satellites (HAPS), usually unmanned airships or airplanes that operate in the stratosphere at 20km, are a promising complementary alternative for GHG and AQ Earth observation applications.

GMV in collaboration with KNMI, ABB and SCEYE, and funded by the European Space Agency developed a project to analyze how HAPS can provide data to operational AQ and GHG services, such as urban AQ modelling or GHG emission inventories. Synergies with existing or planned satellites have also been taken into account.

Key project objectives included:

- The identification of the air quality and GHG modelling user requirements for high-resolution atmospheric composition data to be provided by HAPS, focusing primarily on NO2 emissions, O3, CO2 and particulate matter.

- The demonstration of the impact a HAPS system can have on improving the status of air quality or GHG modelling in synergy with satellite data. Two HAPS use cases were defined, one for the Great Rotterdam region and the other for Seville metropolitan area. Public entities from both regions having the mandate to monitor urban air quality have been involved as end users, providing concrete requirements and needs to identify the existing technical and scientific opportunities and gaps.

- Definition of the mission requirements for the use cases, including the technical platform and the instrument requirements, preliminary system concepts, air space regulations, geophysical data products and synergies with existing and planned satellite missions.

The user requirements collected, discussions with stakeholders and ESA, and a trade-off analysis (balance completeness to fulfil user needs with mission flexibility and cost effectiveness, technological and scientific readiness) led to develop two HAPS mission concepts: Paris Agreement Monitoring Mission and the Metropolitan Surface AQ Mission.

The HAPS Paris Agreement Monitoring Mission would focus on CO2 and NO2. NO2 is a marker fingerprinting CO2 enhancements related to fossil fuel and biomass burning. The wavelength range of the NO2 instrument would be suitable for monitoring of formaldehyde (HCHO) together with NO2. Combination of CO2 with CO and/or CH4 could be made, depending on the instrument design. Although aerosol measurements are not a principal objective of a CO2 emission monitoring mission, auxiliary aerosol measurements to reduce CO2 measurement uncertainty would provide an aerosol product.

The Metropolitan Surface AQ Mission was selected to have a better understanding of the metropolitan ozone pollution and the processes involved, i.e. the emissions of primary pollutants, the influence of local meteorology such as sea breeze in morning and afternoon, and the photochemical interactions. Particulate Matter (PM10; PM2.5; PM1), especially from (agricultural) waste burning is another main source of air pollution. HAPS-based observations could lead to better insights in, e.g., the regional source locations, temporal variability and particle type characterization. This mission would focus on the NO2, tropospheric ozone and aerosol.

The mission concepts were thoroughly analyzed with the objective to define mission configurations with technical solutions for each of the mission technical components: HAPS fleet, instruments payload and ground segment operations. 42 different configurations were deemed of interest and traced against the 100 technical use requirements. The compliance with the user requirements and the analysis of the challenges faced in the different configurations supported the recommendation of the following three configurations: Ready to Fly Mission Configuration, User-Driven Mission Configuration, and Forward-Looking Mission Configuration.
In interaction with the Agency after analysing the mission concepts and configurations, the consortium partners concluded that on balance the most promising solution to explore as a rapidly available technology demonstrator for the mission objective was a demonstration mission primarily focused on NO2 and restricted to the country of Spain:

- NO2 observations would support both air quality regulation and climate emission control policies in Spain as well as provide a demonstration for other European metropolitan regions;

- Spaceborne observations of NO2 are well established and instruments are available with a high TRL, thus optimally combining technological readiness and scientific readiness. Compared to other atmospheric components a much faster user uptake could be foreseen using existing projects, scientific cooperation and other frameworks such as the Copernicus Atmosphere Monitoring Service (CAMS) and its regional spin-offs (high scientific readiness level);

- A demonstration mission restricted to Spain would prevent potential regulatory issues facing a HAPS demonstration mission in Europe crossing national boundaries, e.g. regarding legislation, and aviation safety;

- Compared to the Rotterdam area, aviation safety is much less of an issue over specific areas in Spain such as around Teruel.

Autonomous Surface Vessels (ASVs) offer a unique range of functional, efficiency and safety benefits over traditional manned vessels, by reducing or removing the need for onboard crew. With advanced shipboard autonomy and teleoperation technologies rapidly approaching market-readiness, a range of vessel systems are already undertaking operational demonstrations for applications including bathymetric surveying, naval mine clearance and commercial shipping. The first step towards achieving robust autonomous navigation in complex maritime environments is maintaining an up-to-date Situational Awareness (SA) picture of vessel surroundings, which covers the perception, comprehension and future state prediction of collision hazards, sea conditions, coastal features and other elements comprising the maritime operating environment.

Furthermore, rising rates of armed conflict, piracy and cyber-attacks constitute a significant threat shipping security, trade and supply chains across the globe, raising further concerns for the safety and security of future uncrewed vessel operations. Developing maritime Situational Awareness (SA) capabilities to a level of sophistication and robustness that facilitates safe vessel navigation and collision-avoidance in a variety of high-risk maritime environments is therefore a major enabler of commercially-viable Harbour to Harbour (H2H) autonomous vessel operations.

“Enhanced Surrounding Awareness and Navigation for Autonomous Vessels” (ESANAV) is a recently-concluded technical and commercial system feasibility study led by DEIMOS UK and performed under the ESA Open Space Innovation Platform (OSIP), which seeks to define the route to addressing these challenges using emerging remote sensing technologies, including High-Altitude Pseudo-Satellites (HAPS), next-generation satellite constellations, high-resolution Earth Observation (EO) payloads, state-of-the-art Computer Vision (CV) algorithms and multi-sensor data fusion architectures.

Fixed-wing High-Altitude Pseudo-Satellites are a particular focus of the study, due to their promise of delivering multi-month stratospheric flight endurances and flexible deployment capabilities highly suited to meeting evolving maritime surveillance needs. When equipped with high-resolution optical and Synthetic Aperture Radar (SAR) imaging payloads, these aircraft are foreseen to be capable of 24/7, real-time monitoring with update intervals consistent with the evolution timescale of maritime environments, including under adverse weather conditions where the utility of shipboard sensor suites is significantly degraded.

In addition to enhancing vessel navigation planning and collision avoidance capabilities, the proposed Situational Awareness services would provide significant value to other vessel monitoring entities including Vessel Traffic Services (VTS), maritime security, and regulatory compliance applications. The successful deployment of low-latency (near-)real-time Earth Observation monitoring capabilities will also open up a wide variety of high-value applications in both maritime and non-maritime domains for better understanding human and natural processes on (sub-)hourly timescales, including smart port and smart city asset optimisation, pollution monitoring, conservation efforts, and disaster response coordination.

At LPS22, we will present the final results of ESANAV system architecture definition, technology trade-off and market analysis activities, share the multi-disciplinary 20-year technical and commercial roadmap that constitutes the ultimate outcome of the study, and highlight the foreseen roles that key maritime, UAV/HAPS and space sector stakeholders can play in realising these exciting next-generation Earth Observation systems and services.

MONICAP (MONItoraggio di Colture Agricole Permanente) project, led by Intecs Solutions in cooperation with the Hypatia Research Consortium and CIRA, aims to develop a sustainable system based on a permanent HAPS tethered platform capable of providing very high spatial resolution thermal, multispectral and hyperspectral images (approximately 450nm to 900nm) over an area ranging from 10 to 50 hectares. The solution is a platform composed of an aerostatic balloon, equipped with aerodynamic elements to contribute to overall lift and stabilisation in windy conditions, tied to a ground station. The payload consists of a hyperspectral sensor, a multispectral sensor, a thermal camera and a visible camera, all moved by means of a gimbal that ensures the referenced pointing of each image acquired and the correct management of the hyperspectral sensor. MONICAP provides useful information to support Variable Rate Agriculture applications such as smart fertilisation, smart irrigation, necessity of using herbicide and phytosanitary treatments (such as information for monitoring of downy mildew infestation, esca disease and flavescence dorée). The platform uses an autonomous electrical generation and storage system, through a weight-optimised generator and accumulator, using flexible, ultra-lightweight cells with an efficiency of over 500W/kg and batteries with a capacity of over 300Wh/kg that power all the on-board instrumentation. The significative advantage of MONICAP is due to the permanent acquisition of images, avoiding also the issues related to the cloud coverage, the very high spatial resolution, which is sub-metric for both multispectral/hyperspectral and thermal sensors, the low latency time in data transfer, the possibility of real-time processing, the high payload capacity, the low cost of the service and high security that a tethered platform entails, which makes the whole solution unique compared to satellite systems and drones. Through the processing of vegetation indices (such as NDVI, NDRE, LAI), temperature data, soil moisture and exploiting machine learning/deep learning algorithms, MONICAP is able to produce estimates of crop yield, provide information on nitrogen management, automatically detect plant diseases due to the presence of pests, detect weeds and infesting plants, provide estimates of crop quality, and autonomously recognise different crops species, combining the best of the two worlds of satellite and UAV technologies. MONICAP therefore aims to demonstrate the full potential that HAPS can offer in creating a Digital Twin of our planet, representing a valuable asset for present analysis and future forecasts, and providing an irreplaceable instrument for decision support when integrated with data from satellites, drones and IoT sensors.

SkyRider is HAPS (High Altitude Pseudo Satellite), lighter-than-air platform flying at an altitude of approximately 20km for weeks to months. It can fly missions up to 6 months with payloads up to 10kg and power consumption of 5kW and it can can keep geostationary position in winds up to 15m/s. SkyRider is designed for the main commercial markets such as Earth observation, navigation, telecommunications etc. It is designed for payloads up to 10 kg, managed both vertical and horizontal maneuvering and long-term station-keeping. Station-keeping is especially important for commercial and scientific Earth Observation applications.
Today, the stratosphere is an unjustly neglected layer of the atmosphere in terms of expanding technological infrastructure. We can talk about the so-called "forgotten height," because there is still no human technology or even infrastructure in the stratosphere. At the same time, the stratosphere has the advantage of an ideal height from the point of view for Earth Observation because aircrafts can never reach this height. Furthermore, at this altitude, there is already a permanent access to energy from the Sun, because otherwise the disturbing effects of the weather are associated with the lower troposphere. In addition, thanks to its low air density, the stratosphere allows HAPS to travel quickly on a global scale.
There seems to be a wide range of user commercial demands for services from the stratosphere. At present, there are occasional attempts for the longest possible stratospheric flight or at least a parabolic flight through the stratosphere. The simplest application today is disposable uncontrolled meteorological balloons, which perform measurements in flight (sometimes up to the stratosphere) and transmit data to the Earth by radio probe. It has been discharged for over 50 years several times a day. Leading global technology companies are trying to take a more sophisticated control of the stratosphere. Generally speaking, the rivalry of different concepts focuses on flight length, horizontal and vertical controllability (propulsion), ground segment quality, robust but lightweight energy system, communication links and especially the ability of station-keeping, ie the ability to "hang" in the stratosphere in defined position above the Earth. Obviously, such a concept will be in great demand once it has been designed and successfully tested. The last criterion of competition is also the cost of production and operation of such a system.

An evolution in the Copernicus Space Component (CSC) is foreseen in the second half of the 2020s to meet priority user needs not addressed by the existing infrastructure, and/or to reinforce services by monitoring capability in the thematic domains of CO2, polar, and agriculture/forestry. This evolution will be synergetic with the enhanced continuity of services for the next generation of CSC. Growing expectations about the use of Earth Observation data to support policy-making and monitoring puts increasing pressure on technology to deliver proven and reliable information. Hyperspectral imaging (also known as imaging spectroscopy) today enables the observation and monitoring of surface properties (geo-biophysical and geo-biochemical variables) due to the diagnostic capability of spectroscopy provided through contiguous, gapless spectral measurements from the visible to the shortwave infrared portion of the electromagnetic spectrum. Hyperspectral imaging is a powerful remote sensing technology based on high spectral resolution measurements of light interacting with matter, thus allowing the characterisation and quantification of Earth surface materials. Quantitative variables derived from the observed spectra are diagnostic for a range of new and improved Copernicus services with a focus on the management of natural resources. Thanks to well-established spectroscopic techniques, optical hyperspectral remote sensing has the potential to deliver significant enhancement in quantitative value-added products. This will support the generation of a wide variety of new products and services in the domain of agriculture, food security, raw materials, soils, biodiversity, environmental degradation and hazards, inland and coastal waters, and forestry. These are relevant to various EU policies that are currently not being met or can be substantially improved, but also to the private downstream sector. The Main Mission Objective of the Copernicus Hyperspectral Imaging Mission is: “To provide routine hyperspectral observations through the Copernicus Programme in support of EU- and related policies for the management of natural resources, assets and benefits. This unique visible-to-shortwave infrared spectroscopy based observational capability will in particular support new and enhanced services for food security, agriculture and raw materials. This includes sustainable agricultural and biodiversity management, soil properties characterization, sustainable mining practices and environment preservation.”

The observational requirements of CHIME are driven by the primary application domains i.e. agriculture, soils, food security and raw materials, and are based on state-of-the art technology and results of previous hyperspectral airborne and experimental spaceborne systems. They were drafted by an international group of experts and reflected in the Mission Requirements Document. These baseline observational requirements consider trade-offs and dependencies between parameters such as spectral resolution and radiometric performance.

For the development of the Space Segment Contract (Phase B2/C/D/E1) Thales Alenia Space (France) as Satellite Prime and OHB (Germany) as Instrument Prime were selected. The contract was signed in November 2020 and the corresponding Kick-Off released the start of Phase B2. The System Requirements Review (SRR) was conducted in July 2021 and the Preliminary Design Review (PDR) is planned for mid- 2022. The CHIME Space Segment will be confirmed by the end of the current phase B2. Currently there are 2 satellites foreseen and each of the satellites will embark a hyperspectral instrument, with a single telescope, and three single-channel spectrometers covering each one-third of the total swath of ~130 km. Each spectrometer has then a single detector covering the entire spectral range from 400 to 2500 nm. CHIME will embark a HyperSpectral Instrument (HSI) which is a pushbroom-type grating Imaging Spectrometer with high Signal-to-Noise Ratio (SNR), high radiometric accuracy and data uniformity. The generated Hyperspectral Data are already pre-processed onboard the satellite within the dedicated Data Processing Unit (DPU) allowing cloud detection and compression using artificial intelligence techniques. Once the data are transmitted via Ka-Band antenna to the ground, the Data will be processed and disseminated through the Copernicus core Ground Segment (GS) allowing the generation of CHIME core products: L2A (bottom-of-atmosphere surface reflectance in cartographic geometry), L1C (top-of-atmosphere reflectance in cartographic geometry) and L1B (top-of-atmosphere radiance in sensor geometry).

In this contribution, the main outcomes of the activities carried out in Phase A/B1 and B2, as well as the planned activities for Phase C/D/E will be presented, covering the scientific support studies, the technical developments and the user community preparatory activities. The ongoing international collaboration towards increasing synergies of current and future imaging spectroscopy missions in space will be reported as well.

The Copernicus Imaging Microwave Radiometer (CIMR) expansion mission is designed to provide measurement evidence in support of developing, implementing, and monitoring the impact of the European Integrated Policy for the Arctic. Since the impact of changes in the Polar regions have profound impacts globally, CIMR will provide measurements over the global domain serving users in the Copernicus Ocean, Land, Climate and other Service application domains. The User needs for the CIMR mission are set out in reports from European Commission Polar Expert Group (PEG) user consultation processes, which are supplemented by a document expressing CMEMS recommendations and Copernicus Climate Service User requirements.

The aim of a Copernicus Imaging Microwave Radiometry (CIMR) Mission is to Provide high-spatial resolution microwave imaging radiometry measurements and derived products with global coverage and sub-daily revisit in the polar regions and adjacent seas to address Copernicus user needs. The primary instrument is a conically scanning low-frequency, high spatial resolution multi-channel microwave radiometer. A dawn-dusk orbit has been selected to fly in coordination with MetOp-SG-B1 allowing collocated data from both missions to be obtained in the Polar regions within +/-10 minutes. A conical scanning approach utilising a large 8m diameter deployable mesh reflector with an incidence angle of 55 degrees results in a large swath width of ~2000 km. This approach ensures 95% global coverage each day with a single satellite and no hole at the pole in terms of coverage. Channels centred at L-, C-, X-, Ku- and Ka-band are dual polarised with effective spatial resolution of "≤ 60" km, "≤ 15" km, "≤ 15" km and "< 5" km (both Ka- and Ku-band with a goal of 4 km) respectively. Multiple feeds are used for all but L-band to ensure a slow antenna rotation speed while providing complete coverage of the scanned surface. Projected footprint ellipse on-ground are overlapped and each channel provides 5 samples that are set to ground for each measurement integration time. Measurements are obtained using both a forward scan and a backward scan arc. In-flight calibration is implemented using active cold loads and a hot load complemented by periodic pitch manoeuvres for both deep-space and to the earth surface. On board processing is implemented to provide robustness against radio frequency interference and enables the computation of modified 3rd and 4th Stokes parameters for all channels.

This solution enables a large number of Level-2 geophysical products to be derived over all earth surfaces including sea ice (concentration, thickness, drift, ice type, ice surface temperature) sea surface temperature, sea surface salinity, wind vector over the ocean surface, snow parameters, soil moisture, land surface temperature, vegetation indices, and atmospheric water parameters serving all of the Copernicus Services.

This paper reviews the current status of the CIMR mission, now in Phase B2, the anticipated performance of primary mission Level-2 products that will be provided.

As part of the Copernicus Programme, the European Commission and the European Space Agency (ESA), are expanding the Copernicus Space Component to include measurements for anthropogenic CO2 emission monitoring. The greatest contribution to the increase in atmospheric CO2 comes from emissions from the combustion of fossil fuels and cement production. In support of well-informed policy decisions and for assessing the effectiveness of strategies for CO2 emission reduction, uncertainties associated with current anthropogenic emission estimates at national and regional scales need to be improved.

Satellite measurements of atmospheric CO2, complemented by in-situ measurements and bottom-up inventories, will enable, by using advanced (inverse) modelling capabilities, the transparent and consistent quantitative assessment of CO2 emissions and their trends at the scale of megacities, regions, countries, and at global scale. Such a space capacity, complemented by EUMETSAT development of an operational Ground Segment and data service in place with ECMWF, will provide the European Union with a unique and independent source of information, which can be used to assess the effectiveness of policy measures, and to track their impact towards decarbonising Europe supporting the European Commission’s European Green Deal and meeting national emission reduction targets.

This presentation will provide an overview of the Copernicus CO2 Monitoring (CO2M) mission objectives, the consolidated observational requirements on CO2 and auxiliary measurement capabilities. Operational monitoring of anthropogenic emissions requires high precision CO2 observations (0.7 ppm) with, on average, weekly effective coverage at mid-latitudes. These observations will be obtained from NIR and SWIR radiance spectra at moderate spectral resolution. The measurements will be complemented by (1) aerosol observations, to minimise biases due to incorrect light path corrections, and (2) NO2 observations as tracer for high temperature combustion. Retrieval of CO2 is further facilitated by a cloud imager, to identify measurements contaminated by low clouds and high altitude cirrus. In addition, an update of activities and studies currently undertaken to implement the space component will be presented.

Within the expansion of the Copernicus Sentinel Constellation, the Copernicus Polar Ice and Snow Topography Altimeter (CRISTAL) mission is being developed as a key contribution to Europe’s planned response to the need for monitoring of the polar regions. This need has clearly been identified by an EC-led user consultation process and by the Global Climate Observing System (GCOS). GCOS has recommended continuation of satellite synthetic-aperture radar (SAR) altimeter missions, like the altimeters on board CryoSat-2 and Sentinel-3. CRISTAL will fly to 88° latitude, like CryoSat-2 which is currently in its extended mission phase, therefore ensuring an almost complete coverage of the Arctic Ocean, as well as of the Antarctic ice sheet. CRISTAL will for the first time feature a dual Ku/Ka band SAR altimeter (with interferometric capability on the Ku channel) that enables unprecedented measurements.

The primary objectives of CRISTAL target mainly cryospheric science: measuring and monitoring variability of sea ice thickness and its snow depth, and measuring and monitoring the surface elevation and changes of polar glaciers and ice sheets. CRISTAL will also support applications related to snow cover and permafrost in Arctic regions. In addition to those objectives CRISTAL is expected to contribute significantly to oceanography, like CryoSat-2. CRISTAL will allow observations of global ocean topography up to the polar seas, therefore contributing to global observations of mean sea level, mesoscale and sub-mesoscale currents, wind speed, and significant wave height. This information serves as critical input to operational oceanography and marine forecasting services so it feeds directly into Copernicus’ Marine and Climate Change Services.

In this presentation we will illustrate the advanced technical characteristics of CRISTAL, give an update on its development status (currently in Phase B2) and discuss how this mission extends the heritage of CryoSat-2 over the cryosphere, the oceans and inland waters. We will discuss how the dual-band capability is expected to enable new investigations in the marginal ice zone, in the coastal zone and on surface roughness-related effects, like the sea state bias, and discuss plans for polar campaigns in support of CRISTAL development and cal/val.

The “High Spatio-Temporal Resolution Land Surface Temperature Monitoring (LSTM) Mission” has been identified as one of the Copernicus Expansion Missions. The mission is designed to provide enhanced measurements of land surface temperature in response to presently unfulfilled user requirements related to agricultural monitoring.
High spatio-temporal resolution thermal infrared observations are considered fundamental to the sustainable management of natural resources in the context of agricultural production and with that for global water and food security. Operational land surface temperature (LST) measurements and derived evapotranspiration (ET) are key variables in understanding and responding to climate variability, managing water resources for irrigation and sustainable agricultural production, predicting droughts but also addressing land degradation, natural hazards, coastal and inland water management as well as urban heat island issues. Earth observation (EO) monitoring products based on thermal observations, are therefore considered important for informed policy making, including amongst others the UN Sustainable Development Goals (e.g. SDG 6.4), the UN Convention for Combating Desertification and Land Degradation, the UN Water Strategy, the EU Common Agriculture Policy, the EU Policy Framework on Food Security, the EU Water Framework Directive, the EU 2030 agenda for Sustainable Development and the recent EU Green Deal ambitions.
The existing Copernicus space infrastructure, including in particular the Sentinel-1 and Sentinel-2 missions, already provides useful information for agricultural applications. Although Sentinel-3 routinely delivers global LST measurements, its limited 1 km spatial resolution does not capture the field-scale variability required for irrigation management, crop growth modelling and reporting on crop water productivity. In view of the foreseen evolution of the Copernicus program, additional high-level observation requirements have been collected by the European Commission as part of a user survey and further assessed at the Copernicus Agriculture and Forestry User Requirement Workshop in 2016, revealing the lack of European spaceborne capability for providing high spatio-temporal resolution Thermal Infrared (TIR) observations . Therefore, a dedicated LSTM mission is foreseen in the frame of the Copernicus expansion with the following mission objectives:
• Primary objective: to enable monitoring evapotranspiration rate at European field scale by capturing the variability of LST (and hence ET) allowing more robust estimates of field scale water productivity
• Complementary objective: to support the mapping and monitoring of a range of additional services benefitting from TIR observations – in particular soil composition, urban heat islands, coastal zone management and High-Temperature Events.

The LSTM mission will deploy two satellites equipped with TIR instruments optimised to support agriculture management services with the specific mission objectives above. In response to the priority user needs, the Mission Requirements Document (MRD) for the space component has been developed by an international Mission Advisory Group under European Space Agency (ESA) leadership . The key observational requirements of the LSTM mission, as outlined in the MRD, are systematic global acquisitions of high-resolution (50 meters) observations with a high revisit frequency (1-3 days) in 3-5 thermal bands (8-12.5 m) accompanied with a number of VNIR-SWIR spectral bands. The accuracy for LST measurements shall be better than 1-1.5 K at a 300 K reference temperature. The MRD serves as input for the mission design, by conveying the EU Policy framework, the user needs, the mission objectives and the observation requirements for each Copernicus candidate mission.
ESA is collaborating with partner space agencies to create synergy with relevant international missions such as TRISHNA (CNES, ISRO), Surface Biology Geology SBG (NASA/JPL) and the Landsat program (USGS/NASA) with the aim to achieve the optimal temporal coverage of high-resolution thermal observations.
This presentation will provide an overview of the proposed Copernicus LSTM mission including the user requirements, a technical system concept overview, Level-1/Level-2 core products description and a range of use cases addressing the mission objectives. The LSTM mission has started its phase B2 and successfully placed a contract in late 2020 with an industrial consortium led by Airbus Spain. In spring 2021 the mission successfully passed the System Requirements Review.

1 Agriculture & Forestry Applications User Requirements Workshop Report (2016): http://workshop.copernicus.eu/sites/default/files/content/attachments/form-WfbHTJJLH6suSlxf09G4p6pXsUAIEArRc76DBmZ3lDA/agri_forestry_ws_final_report.pdf
2 LSTM Mission Requirements Document, version 3: https://www.esa.int/Applications/Observing_the_Earth/Copernicus/Copernicus_High_Priority_Candidates

The Radar Observing System for Europe at L-band (ROSE-L) is part of the Copernicus Expansion Programme which focuses on new missions that have been identified by the European Commission (EC) as priorities for implementation in the coming years. ROSE-L provides additional capabilities above and beyond those of the current Sentinel missions, filling observation gaps as well new and emerging user needs not yet addressed.
ROSE-L is a user driven mission. By filling important observation gaps in the current Copernicus satellite constellation, the ROSE-L mission supports key European policy objectives and provides enhanced continuity for a number of Copernicus services and down-stream commercial and institutional users. Due to the longer wavelength, L-Band SAR observations from space provide additional information that cannot be gathered by other means benefiting a variety of services and applications.

A high-level mapping between specific European policy objectives and the unique information provided by the mission is provided below. The mission will contribute inter alia to:
- The safety of European Citizens by greatly extending the monitoring of geohazards linked with surface motion such as landslides, subsidence and earthquake/volcanic phenomena into vegetated areas which are inaccessible to current Copernicus satellites and will be critical to the nascent European Ground Motion Service (EU-GMS);
- The European Arctic policy and the sustainable economic development of the Arctic region by providing new information sea ice types and detection of icebergs critical to safe navigation and building of infrastructure in Arctic areas;
- Forestry and maintaining biodiversity through the continuous high-resolution monitoring of changes in global forest carbon stocks and their spatial distribution;
- Agriculture and food security by providing reliable high-resolution soil moisture information to support improved management of water use, enhances weather-independent land cover and crop information, feeding meteorological and hydrological forecast models;
- EU Water Framework Directive through mapping of water availability and water use particularly for agriculture;
- Climate change policy through the enhanced monitoring of glaciers and ice sheets, forest carbon stocks and changes with time and water availability;
- The European Union Integrated Maritime Policy by extending the capacity to monitor our marine ecosystem and by increasing our maritime surveillance abilities.

In terms of user level information products the ROSE-L mission includes the following
- Line-of-sight surface motion addressing deformation measurements, urban subsidence, landslides, flooding
- Forest above ground biomass, forest area, forest change, land cover maps, crop type and status products to support Land use, Land use change, forestry and agriculture
- Soil moisture at regional and global scale to support improved weather forecasts, hydrology and water management
- Sea ice type, sea ice concentration, sea ice motion, glacier/ice cap surface velocity, grounding line and snow water equivalent (SWE) in support of Cryosphere and Arctic application needs
- Wind and wave spectrum information over oceans for regular forecast, EMR and extreme events
- Vessel detection oil spill mapping and ice berg detection in support of maritime security

The requirements and implementation of the ROSE-L SAR mission cannot be considered in isolation but need to build on existing and planned Copernicus observation capabilities and new commercial/NewSpace SAR developments to derive maximum benefits for users and services. Observational gaps filled by ROSE-L requires careful combination of the new information with information provided by existing Sentinel missions. The enhanced continuity also requires harmonised, coordinated and systematic acquisitions in conjunction with other Sentinel data, and in particular those provided by the C-band radar aboard Sentinel-1.

The ROSE-L SAR instrument will operate in L-band, i.e. in the frequency range from 1.215 to 1.300 GHz. The L-band SAR instrument will be able to operate in SAR modes suitable for imaging land and coastal areas, as well as sea-ice and open ocean. Derived from the high-level mission objectives, the SAR instrument of the ROSE-L mission is currently based on three main imaging modes:
- Dual-polarisation (co- and cross-polarisation)
- Fully-polarimetric
- Wave Mode.

The dual- polarisation mode represents the nominal “work-horse” imaging mode of ROSE-L, to enable systematic imaging over global land and ice, with a focus on Europe. It combines a large swath with high resolution and high quality imaging specifications e.g. -28 dB NESZ. This mode meets most user requirements in the various application areas supported by the ROSE-L mission. By selecting a main mode of operation, conflicting requests from users and corresponding gaps in acquisitions are avoided and a consistent and complete archive of data to support long-term assessments of trends is secured.

ROSE-L is implemented as a 3-axis stabilized satellite based on the new Thales Alenia Space Multi-Mission Platform product line (MILA) and will embark the L-Band Synthetic Aperture Radar (SAR) Instrument dedicated to the day-and-night monitoring of land, ice and oceans offering improved revisit time, full polarimetry, high spatial resolution, high sensitivity, low ambiguity ratios and capability for repeat-pass and single-pass cross-track interferometry. It is based on a 5-panel deployable 11 m × 3.6 m L-Band, highly innovative and lightweight planar Phased Array Antenna (PAA). The satellite will also carry a set of three Monitoring Cameras (CAM) to monitor the deployment of the SAR antenna and the solar arrays.

To support both the climate modelling and carbon cycle science communities, the ESA Climate Change Initiative (CCI) Biomass project is producing maps depicting the global distribution of woody above-ground biomass at 100 m spatial resolution as well as the uncertainty (standard deviation) of the derived estimates. In the first three years of the project, maps have been produced for the years 2010, 2017, and 2018 based on advanced versions of the retrieval algorithms developed in the frame of the predecessor ESA GlobBiomass project (Santoro et al., 2021). The project relies on multi-temporal stacks of spaceborne C- and L-band radar data acquired by the ESA C-band SAR missions ENVISAT ASAR (Wide-Swath mode) and Sentinel-1 (Interferometric Wide-Swath mode) and JAXA’s L-band SAR missions ALOS-1 PALSAR (Fine Beam Dual-Polarization mode) and ALOS-2 PALSAR-2 (Fine Beam Dual-Polarization and ScanSAR modes) for mapping above-ground biomass. In addition, the mapping of above-ground biomass considers spaceborne LiDAR (ICESAT GLAS) data to support the modeling of multi-temporal C- and L-band radar backscatter with information on forest structural differences reflected in varying allometric relationships between forest height, density, and above-ground biomass. An independent validation based on in situ plots distributed across the major forest biomes, albeit not with a systematic sampling design, as well as intercomparisons with airborne LiDAR-derived biomass maps available for various sites in South and North America, Africa, Europe, Southeast Asia, and Australia confirmed that the CCI Biomass products are of better quality than GlobBiomass but still have regional biases and a per-pixel uncertainty of about 30-40%.

The quantification of annual and decadal changes in above-ground biomass is a critical component of CCI Biomass. Based on the above-ground biomass for three different years, CCI Biomass has released change products for the years 2010-2018 and 2017-2018. The change maps are accompanied by quality flags indicating the reliability/probability of the reported changes. However, quantification of pixel-level changes beyond those flagged as probable in the released change products is currently discouraged. Intercomparisons of the three biomass maps and associated uncertainties highlighted the limitations to the estimation of biomass on a global scale, which could be categorized as signal- or processing-dependent. The signal-dependent limitations relate to the varying sensitivity of C- and L-band backscatter to biomass as well as the insufficient characterization of forest structure in the modeling locally. Processing-dependent limitations were a consequence of local or systematic imperfections in the pre-processing of the available radar data to radiometrically terrain-corrected level. Furthermore, radar data acquired by different satellite missions had to be used for the three different epochs and this posed restrictions on the inter-annual harmonization of global maps. This was largely attributed to the different acquisition modes and inconsistent multi-temporal acquisition plans, resulting in a strongly varying number of observations as well as seasonal coverage between years.

Future activities in CCI Biomass will seek to improve the inter-annual consistency of above-ground biomass estimates when reproducing the maps for the years 2010, 2017, and 2018 as well as producing maps for additional years between 2015 and 2022. Continued efforts will be made on optimizing the retrieval algorithms considering new spaceborne LiDAR data acquired by GEDI and ICESAT-2 as well as additional field data. In addition, through an ESA-JAXA cooperation on biomass estimation, JAXA will make available to the CCI Biomass consortium an exclusive dataset of ALOS-1 PALSAR and ALOS-2 PALSAR-2 imagery that will be reprocessed with quality superior to the publicly available data mosaics available so far.

References
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

Accurate estimation of aboveground forest biomass stocks is required to assess the impacts of land use changes such as deforestation and subsequent regrowth on concentrations of atmospheric CO2. The Global Ecosystem Dynamics Investigation (GEDI) is a lidar mission launched by NASA to the International Space Station in 2018 and has now completed 3 years of canopy structure observations required for biomass estimation. GEDI was specifically designed to retrieve vegetation structure within a novel, theoretical sampling design that explicitly quantifies biomass and its uncertainty across a variety of spatial scales. Here we report on GEDI’s approach to biomass estimation and its resulting estimates of pan-tropical and temperate biomass that span areas from 25 m footprints to mean values and uncertainties at sub-national and country levels. We begin with an overview of the GEDI mission and provide details of its technical and mission implementation, including its lidar instrument. We then present a summary of GEDI data products, including canopy metrics, with respect to their quality and quantity. We next briefly describe GEDI’s statistical framework and illustrate the process by which GEDI was designed to support its use. We provide the most current biomass results from the mission and provide comparisons with national forest inventory data from the United States and for other countries. Finally, we address the assumptions and limitations of GEDI’s approach and consider areas where improvements may be warranted. The results reported here represent a watershed product of the first space mission longitudinally coordinated, from engineering to estimation, to generate biomass products in a transparent way with errors that are well-characterized using established probability theory. The GEDI investigation highlights the great value of an approach that explicitly address uncertainty as integral part of mission design and suggests that future space missions should carefully consider adopting a similar strategy, as appropriate.

Forests play a critical role in the global carbon cycle, storing approximately 500 Pg of aboveground biomass (Santoro et al. 2021), and forest loss contributes approximately 14% to atmospheric warming. Forest management is an important avenue for climate mitigation, both through avoiding emissions related to deforestation and degradation, and bolstering carbon sinks from afforestation and regrowth. The carbon estimates associated with forest losses and gains are highly uncertain, largely because of a lack of reliable forest carbon stock (aboveground biomass) maps. Indeed, past estimates of forest aboveground biomass stocks and fluxes vary greatly both within and between forest systems (Spawn et al. 2020). Accurately mapping forest biomass at a global-scale is a priority for a suite of new and upcoming satellite missions from NASA, ESA and JAXA, including GEDI, ICESat-2 ALOS-2, ALOS-4, NISAR and BIOMASS. The first set of new (circa 2020) biomass products have recently become available (https://earthdata.nasa.gov/maap-biomass/), many using inputs from this suite of new satellite instruments. While these products should represent increased accuracies in comparison to pre-2020 products, their relative accuracies have yet to be assessed across the range of biomes they represent. Indeed, discrepancies between products may reduce their uptake and confuse users, and harmonizing these products for policy applications (e.g. the UNFCCC’s Global Stocktake) is highly desirable. Transparent product validation and inter-comparison is critical to facilitate the improvement and uptake of these new biomass products, and other products that come online in future.

A recent international collaborative effort between biomass map producers and users organized under the Committee for Earth Observation Satellites (CEOS) Agriculture Forestry and Other Land User (AFOLU) seeks to fulfill this need. This activity, hosted on the NASA-ESA Multi-Mission Algorithm and Analysis Platform (MAAP), is an Open Science activity aimed at increasing transparency and collaboration for biomass mapping and validation. Here we present early intercomparison and validation results for 2020 biomass products. We include an intercomparison of biomass maps at a biome-by-biome scale (e.g. Moist Tropical Forests, Mangroves, Boreal Forests) to better understand discrepancies between new products. Additionally, independent reference data from airborne lidar biomass maps are used for pixel-level validation of products with samples across a subset of biomes in tropical, temperate and boreal systems. A second approach to validation using the novel plot2map tool provides insights into product performance at the policy-relevant, national and jurisdictional-scale for which harmonization and targeted estimation are planned. These analyses further inform data users on which products may be most suitable for their area of interest, and will be subsequently used to a) improve EO products and b) guide product harmonization at policy-relevant scales. This presentation includes the latest updates from the CEOS biomass harmonization team and links to a second abstract by Melo et al. focused on the importance of engagement with countries in the harmonization process.

The role of forest biomass in the European bioeconomy is of increasing importance. The assessment of the current availability and the modelling of the potential supply of forest biomass require harmonized statistics and maps indicating how much biomass is available for wood supply and its increment.

In the European context, the biomass assessment is traditionally performed by the National Forest Inventories (NFIs) using extensive field sampling. Yet, satellite and airborne Earth Observation (EO) data are increasingly being used to spatially integrate and intensify the monitoring frequency of ground-based data by mapping forest properties over large areas. However, as European NFIs employ country-specific forest and forest biomass definitions and estimation methods, and since related estimates refer to different periods and spatial scales, it is essential to harmonize the ground-based biomass statistics and EO-based maps to perform a meaningful pan-European biomass assessment. To support this goal, we present a comprehensive study that includes the following components towards a full harmonization and integration of maps and plot-based statistics.

First, the biomass statistics from NFI assessments were harmonized in terms of biomass definition and statistical estimator for expansion from plot level to regional estimates, through collaboration of 26 European NFIs. The biomass regional estimates were then further harmonized to a common reference year using the Carbon Budget Model, a forest growth model developed by the Canadian Forest Service and adapted to the specific European conditions.

Second, the resulting biomass statistics was used as a reference to assess the uncertainties of EO-based biomass maps. The map with the highest accuracy was selected and was then modified with a bias-removal correction, removing the observed systematic difference of this map with the harmonized statistics. The resulting 1-ha forest biomass map for Europe is in line with the reference statistics in terms of forest area and biomass stock.

Third, the national statistics of 22 European NFIs were also harmonized for the Forest area Available for Wood Supply (FAWS) and related biomass stocks, using the same reference definition and common criteria to assess wood availability and related restrictions. These harmonized statistics were used to map the FAWS in Europe using environmental and economic spatially-explicit restrictions. The mapped restrictions considered the following parameters and related datasets: (i) forest accessibility, excluding areas above a certain altitude and slope (derived from the Copernicus EU-DEM) and distance to roads (derived from the Open Street Map database); (ii) the legal restrictions to the forest use, excluding the protected areas with no or minimal management (IUCN reserves and national parks as mapped in the World Database on Protected Areas) and protected tree species (according to their probability of presence derived from the JRC European Atlas of Forest Tree Species); and (iii) the areas where the forest productivity (estimated using the novel kNDVI vegetation index computed from MODIS NDVI 250 m data) is deemed to be too low for sustainable timber extraction. The thresholds of each restriction were adjusted according to the country circumstances to account for the differences in the forestry sectors. The FAWS map was then applied to the harmonized forest biomass map to identify the biomass available for wood supply in Europe.

Lastly, the assessment of the available biomass stock is complemented with accurate and comparable estimates of the biomass increment, which are essential to quantify the sustainable supply of biomass from the forest sector. For this purpose, a dedicated study on forest increment is currently being performed by 11 European NFIs to provide harmonized estimates of the forest gross and net annual increment. The preliminary results of this study are presented and compared with existing satellite-based products related to forest productivity to investigate the coherence between the EO maps and plot-based statistics related to biomass growth.

Above-ground biomass (AGB) and its change (∆AGB) are essential variables for dynamic global climate models and national reporting of carbon profiles (Herold et al., 2019). While there is an increasing availability of space-based estimates of AGB in multiple periods, existing AGB maps cannot be simply subtracted to obtain ∆AGB values, as owing to uncertainties, the map differences are not depicting true changes. The issue is illustrated by the disagreements among different map-based ∆AGB estimates (Figure 1). Here we provide an assessment and inter-comparison of global forest AGB change products (2018-2010) from recent publicly released sources: the European Space Agency Climate Change Initiative (ESA-CCI) 100-m AGB maps version 3 (Santoro and Cartus 2021); the World Resource Institute 2000-2020 Carbon Flux Model (WRI-Flux) (Harris et al., 2021) that we modified to produce 2018-2010 AGB fluxes at 30-m pixel size; and the 10-km “JPL” global time series AGB (Xu et al., 2021). For each map-based product, we also produced bias-adjusted counterparts following our uncertainty assessment framework that includes bias prediction as a function of spatial covariates (Araza et al., under review). The assessments were done at 10-km aggregation level for all products and at finer spatial resolution (500 m and 1 km) without the JPL product. Several independent reference data with uncertainty estimates were used to evaluate the map-based ∆AGB consisting of re-measured National Forest Inventory (NFI) and periodic high-resolution AGB maps from airborne LiDAR (local level) and satellite images (regional level). The reference datasets were from ten countries within the four major ecological zones. The ∆AGB estimates were further compared against the Forest Resource Assessment (FRA) country data.

The preliminary results revealed that the assessment of map-based AGB losses and gains depends on the aggregation scale and the choice of reference data. At 1-km scale, map-based AGB losses and gains compare well with the LiDAR data and they are slightly underestimated when compared with NFIs and regional maps. The underestimation of AGB losses is evident to all reference data at 10-km comparisons especially when using NFI and at the country-level regardless of the FRA reporting capacity. The underestimated AGB gains are lessened at such coarser levels except when using NFI. This slight improvement of map-based AGB gains at coarser levels is also the added value of the bias adjustment. Moreover, correcting for map biases also helps reduce the map disagreements particularly in dry Africa woodlands and boreal regions in Figure 1. Outcomes of the map assessments will be the basis to use any individual product or a harmonized product for national carbon accounting in pilot countries.


*******INSERT FIGURE 1********

Figure 1. Overlap of ∆AGB (loss as < -10 Mg/ha; gain as > 10 Mg/ha; no change as < 10 to > -10 Mg/ha) among the three global AGB change products (ESA-CCI, WRI-Flux, JPL) epoch 2018-2010 at 0.1° spatial resolution and without bias adjustment. The map classes portray whether the 3 products agree/disagree or only 2 of them agree/disagree. An example where all products disagree is when a certain pixel depicts: loss (ESA-CCI), gain (JPL), and no change (WRI-Flux).



 
References:


Araza, A. et al. (2021). A comprehensive framework for assessing the accuracy and uncertainty of global 375 above-ground biomass maps. (Manuscript under review)

Harris, N. L., Gibbs, D. A., Baccini, A., Birdsey, R. A., De Bruin, S., Farina, M., ... & Tyukavina, A. (2021). Global maps of twenty-first century forest carbon fluxes. Nature Climate Change, 11(3), 234-240.

Herold, M., Carter, S., Avitabile, V., Espejo, A. B., Jonckheere, I., Lucas, R., ... & De Sy, V. (2019). The role and need for space-based forest biomass-related measurements in environmental management and policy. Surveys in Geophysics, 40(4), 757-778.

Santoro, M.; Cartus, O. (2021): ESA Biomass Climate Change Initiative (Biomass_cci): Global datasets of forest above-ground biomass for the years 2010, 2017 and 2018, v3. NERC EDS Centre for Environmental Data Analysis.

Xu, L., Saatchi, S. S., Yang, Y., Yu, Y., Pongratz, J., Bloom, A. A., ... & Schimel, D. (2021). Changes in global terrestrial live biomass over the 21st century. Science Advances, 7(27), eabe9829.

The forests and savannahs of Africa are amongst the most pristine and biodiverse ecosystems on Earth and collectively contain large carbon stocks in the form of biomass. Despite its global importance, the African continent is one of the weakest links in our understanding of the global carbon cycle due to its sparse observation network. The CarboAfrica project estimated that the biogenic carbon balance of sub-Saharan Africa is currently a net sink of between 0.16 and 1.00 Pg C yr-1 [1]. However, other studies indicated a very small sink or a near to neutral balance [2,3], but that this may be declining or already transitioning into a source [3-5]. In contrast, process-based model estimates present a far larger and unrealistic sink of 3.23 Pg C yr-1 (ranging from 1.3 - 3.9 Pg C yr-1) [1].

In this study we analysed continent wide aboveground woody biomass (AGB) dynamics using a time series of AGB maps for 2007 to 2017. We developed these maps at a spatial resolution of 100m using Global Ecosystem Dynamics Investigation (GEDI) LiDAR footprints [6], Airborne Laser Scanner (ALS)-based AGB maps, temporally cross-calibrated Synthetic Aperture Radar (SAR) ALOS PALSAR / ALOS-2 PALSAR-2 mosaics [7], and Landsat Percent Tree Cover [8]. Our approach consisted of a Random Forests Regression algorithm within a spatial k-fold cross-validation framework, followed by empirical modelling to generate AGB predictions and uncertainty outputs as in Rodríguez-Veiga et al. [9]. We validated our AGB maps with a large dataset of reference field data distributed across the continent (circa 11,000 field plots). Our results show that the AGB stocks in Africa were approximately 120.5 Pg during the study period. When estimating AGB gains and losses in Africa, we observe that the AGB inter-annual stock changes are nearly zero at the beginning of the period, but a continuous increase in the annual rate of deforestation, especially in the Congo Basin, is driving a negative trend in inter-annual AGB stock changes in recent years.



References:
1. Bombelli, A.; Henry, M.; Castaldi, S.; Adu-Bredu, S.; Arneth, A.; Grandcourt, A.d.; Grieco, E.; Kutsch, W.L.; Lehsten, V.; Rasile, A. An outlook on the Sub-Saharan Africa carbon balance. Biogeosciences 2009, 6, 2193-2205.
2. Ciais, P.; Piao, S.L.; Cadule, P.; Friedlingstein, P.; Chédin, A. Variability and recent trends in the African terrestrial carbon balance. Biogeosciences 2009, 6, 1935-1948, doi:10.5194/bg-6-1935-2009.
3. Williams, C.A.; Hanan, N.P.; Neff, J.C.; Scholes, R.J.; Berry, J.A.; Denning, A.S.; Baker, D.F. Africa and the global carbon cycle. Carbon balance and management 2007, 2, 1-13.
4. Hubau, W.; Lewis, S.L.; Phillips, O.L.; Affum-Baffoe, K.; Beeckman, H.; Cuní-Sanchez, A.; Daniels, A.K.; Ewango, C.E.; Fauset, S.; Mukinzi, J.M. Asynchronous carbon sink saturation in African and Amazonian tropical forests. Nature 2020, 579, 80-87.
5. Baccini, A.; Walker, W.; Carvalho, L.; Farina, M.; Sulla-Menashe, D.; Houghton, R.A. Tropical forests are a net carbon source based on aboveground measurements of gain and loss. Science 2017, 10.1126/science.aam5962, doi:10.1126/science.aam5962.
6. Dubayah, R.; Blair, J.B.; Goetz, S.; Fatoyinbo, L.; Hansen, M.; Healey, S.; Hofton, M.; Hurtt, G.; Kellner, J.; Luthcke, S. The Global Ecosystem Dynamics Investigation: High-resolution laser ranging of the Earth’s forests and topography. Science of remote sensing 2020, 1, 100002.
7. Shimada, M.; Ohtaki, T. Generating Large-Scale High-Quality SAR Mosaic Datasets: Application to PALSAR Data for Global Monitoring. Selected Topics in Applied Earth Observations and Remote Sensing, IEEE Journal of 2010, 3, 637-656.
8. 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., et al. High-Resolution Global Maps of 21st-Century Forest Cover Change. Science 2013, 342, 850-853, doi:10.1126/science.1244693.
9. Rodríguez-Veiga, P.; Carreiras, J.; Smallman, T.L.; Exbrayat, J.-F.; Ndambiri, J.; Mutwiri, F.; Nyasaka, D.; Quegan, S.; Williams, M.; Balzter, H. Carbon Stocks and Fluxes in Kenyan Forests and Wooded Grasslands Derived from Earth Observation and Model-Data Fusion. Remote Sensing 2020, 12, 2380.

1. Abstract

As many other research fields, remote sensing has been greatly impacted by machine and deep learning and benefits from technological and computational advances. In the recent years, a considerable effort has been spent on deriving not just accurate, but also reliable modeling techniques. In the particular framework of image classification, this reliability is validated by e.g. checking whether the confidence in the model prediction adequately describes the true certainty of the model when confronted with unseen data. We investigate this reliability in the framework of classifying satellite images into different land cover classes. More precisely, we use the So2Sat LCZ42 data set [1] comprised of Sentinel-1 and Sentinel-2 image pairs. Those were classified into 17 categories by a team of two labelers, following the Local Climate Zone (LCZ) classification scheme.

As a novelty, we make explicit use of the so-termed evaluation set which was additionally produced by the authors of the LCZ42 data set. In this supplementary study, a subset of the initial data was re-labeled by 10 different remote sensing experts, which independently of one another re-cast their label votes for each satellite image. The resulting sets of label votes contain a notion of human uncertainty associated with the underlying satellite images. In the following, we try to explicitly incorporate this uncertainty into the training process of a neural network classifier and investigate its impact on model performance. Also, the earlier introduced definition of reliability is checked and compared to a more common modeling approach. The more common approach is using a single ground truth as label, which is derived from the majority vote of the individual expert label votes.

2. Methodology

The 17 LCZs describe the urbanization of certain cities and are comprised of 10 classes related to built-up areas (urban classes) and 7 classes related to surrounding land cover (non-urban classes). The evaluation data set, which we will use for modeling purposes in the following, consists of 10 European cities as well as additional areas from around the globe which are added for class balancing reasons. A total of ca. 250.000 Sentinel-1 and Sentinel-2 image pairs are included, with a total of 10 spectral bands and 8 statistics derived from the VV-VH dual Pol SLC Sentinel-1 data. Each image is of size 32 by 32 pixels and covers an area of 320m by 320m. For simplicity, we will focus our analysis only on the Sentinel-2 data. Accompanying each satellite image, 10 individual expert label votes are provided. These votes are aggregated for each image by forming the empirical distribution over the different classes. As a result, we receive a distributional label form that stores the information from the individual label votes. Additionally, we store the majority vote of the experts for each image, which serves as a pseudo ground truth label. In case of a tie, the label cast by the two initial labelers was also considered for the determination of the majority vote. Due to the overall high rate of agreement among the voters within the non-urban classes, solely the images associated with the urban classes are considered for modeling the distributional labels in the following.

As a result, the human uncertainty is now saved within the derived label distribution. For the purpose of integrating this information into the classification task, two main changes are made to an already existing deep neural network. First, the usual one-hot encoded labels are replaced by the computed distributional labels. Furthermore, the typical cross-entropy loss gets replaced by the Kullback-Leibler (KL) divergence. This is done to better reflect the task of approximating the ground truth distribution, which is formed by the label votes, from an information theoretic perspective. The training is performed as usual by backpropagating the loss through the network. For evaluating the predictive uncertainty of the model, we investigate the so termed expected calibration error (ECE). The ECE is derived from comparing the model confidence (i.e. the highest predicted class probability of the model) and the corresponding accuracy on the hold-out test set. The discrepancies between the two quantities can further be visualized in a 2D bar plot called reliability diagram.

3. Experiments & Results

We use the benchmark model for the data set from a previous study [2], in which the authors found this model to be superior over many common Convolutional Neural Network (CNN)-based architectures. This benchmark model is termed Sen2LCZ, and builds on the combination of conventional convolutional blocks, the fusion of multiple intermediate deep features and double-pooling. Our implementation results in a network depth of 17 and uses a dropout after the second and third block. The evaluation data set was split into geographically separated training and testing sets. The latter was furthermore randomly split into validation and testing data.
Two separate implementations of the benchmark model were evaluated in order to identify the impact of explicitly modeling the human uncertainty in the labels. The classical approach employed the one-hot encoded labels based on the majority vote of the label votes together with the typically used cross-entropy loss. On the other hand, the modified model utilized the earlier described distributional labels as well as the KL divergence as loss. Apart from that, identical architectures, hyperparameters and training setups were applied. The usual performance metrics were derived on the identical unique test set for both implementations.

As a first and foremost result, all metrics including overall accuracy, average accuracy (both macro and weighted) as well as the kappa score, improved by at least 1 percentage point when using the distributional labels. Note that for deriving these metrics in the presence of distributional labels, the majority vote (i.e. the mode of the distributional label) was taken as ground truth, and the prediction was counted as correct if this ground truth was matched by the highest predicted class probability. Furthermore, the cross-entropy between the predicted probabilities and the ground truth one-hot labels could be reduced by ca. 20% on the test set by training with distributional labels. The central result of this work can be moreover seen in the accompanying visualization, which shows the reliability diagrams of the two implementations: The expected calibration error could be reduced by a large margin (cut more than half) via incorporating the label distributions, and overconfidence could be avoided. The average confidence matches the overall accuracy, and furthermore the two quantities are closely related for almost the entire spectrum.

4. Conclusion
The last reported result shows the clear advantage of integrating label uncertainty into the training process of a neural network for the task of classifying satellite images into LCZs. Adding to that, the integration is superior over classical calibration methods as it also led to improved model performance metrics and a reduced loss on the test set. The derivation and implementation of the distributional labels are straightforward and easy to use. As a main outcome, we would like to emphasize the large improvement in the calibration of the predictive distribution. In particular, the predicted probabilities of the model using the distributional labels can be solidly interpreted and adequately reflect the uncertainty in the prediction.

References:
[1] Zhu, X. X., Hu, J., Qiu, C., Shi, Y., Kang, J., Mou, L., Bagheri, H., Hua, Y., Huang, R., Hughes, L.H., Li, H., Sun, Y., Zhang, G., Han, S., Schmitt, M., Wang, Y. (2020). So2Sat LCZ42: A benchmark data set for the classification of global local climate zones. IEEE Geoscience and Remote Sensing Magazine (GRSM), 8(3), 76-89.
[2] Qiu, C., Tong, X., Schmitt, M., Bechtel, B., & Zhu, X. X. (2020). Multilevel feature fusion-based CNN for local climate zone classification from sentinel-2 images: Benchmark results on the So2Sat LCZ42 dataset. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 13, 2793-2806.

Understanding regional carbon dioxide (CO₂) surface fluxes is an important problem in climate science. To estimate these surface fluxes the usual approaches are based on inverse modeling using atmospheric CO₂ observations. In addition to CO₂ measurements, other gases have shown to be linked to CO₂ fluxes. Nitrogen dioxide (NO₂) has been used as a proxy for anthropogenic CO₂ emissions, both at regional scale (Hakkarainen et al., 2016; Reuter et al., 2014) as well as at individual power plant or city level (Hakkarainen et al., 2021; Reuter et al., 2019). Solar-induced fluorescence (SIF) also plays a role as indicator of vegetation gross primary production. Carbon monoxide (CO) is connected with biomass-burning emissions (Lin et al., 2020).

In this work we take a machine learning (ML) approach to predict global monthly fluxes of CO₂ based on satellite observations of CO₂ and SIF from NASA's Orbiting Carbon Observatory-2, OCO-2, NO₂ observations from the OMI instrument on board of NASA's Aura satellite and CO observations from MOPITT/Terra. We do not use the geographic location of these observations in our model, as we are interested in a model independent of the location. As training data for CO₂ fluxes, we use monthly estimates from CarbonTracker CT2019b. We focus on the years 2015–2021, as OCO-2 was launched in 2014. Since current CarbonTracker CT2019b global estimates go until December 2018, we use observations from 2015 to 2017 as training data and 2018 measurements as test data.

After comparing different ML regression models, we conclude that the best option is to use a XGBoost model, which is the one with lowest mean absolute error. We show that the monthly CO₂ fluxes predicted by our model agree with those derived by the CarbonTracker CT2019b, with small differences in certain areas and months. Our results indicate that NO₂ measurements play the most important role to derive CO₂ fluxes, followed by SIF observations. This supports the importance of NO₂ to detect anthropogenic CO₂ emissions. We make further predictions for the years 2019 to 2021 and detect the reduction of CO₂ emissions due to the effect of COVID-19 lockdowns in 2020.

References

Hakkarainen, J., Ialongo, I., Tamminen, J., 2016. Direct space-based observations of anthropogenic CO₂ emission areas from OCO-2. Geophysical Research Letters 43, 11,400–11,406. doi:https://doi.org/10.1002/2016GL070885.

Hakkarainen, J., Szeląg, M.E., Ialongo, I., Retscher, C., Oda, T., Crisp, D., 2021. Analyzing nitrogen oxides to carbon dioxide emission ratios from space: A case study of Matimba Power Station in South Africa. Atmospheric Environment: X 10, 100110. doi:https://doi.org/10.101/j.aeaoa.2021.100110.

Reuter, M., Buchwitz, M., Hilboll, A., Richter, A., Schneising, O., Hilker, M.,Heymann, J., Bovensmann, H., Burrows, J.P., 2014. Decreasing emissions of NOx relative to CO₂ in East Asia inferred from satellite observations. Nature Geoscience 7, 792–795. URL: https://doi.org/10.1038/ngeo2257, doi:10.1038/ngeo2257.

Reuter, M., Buchwitz, M., Schneising, O., Krautwurst, S., O’Dell, C.W., Richter, A., Bovensmann, H., Burrows, J.P., 2019. Towards monitoring localized CO₂ emissions from space: co-located regional CO₂ and NO₂ enhancements observed by the OCO-2 and S5P satellites. Atmospheric Chemistry and Physics 19, 9371–9383. URL: https://acp.copernicus.org/articles/19/9371/2019/, doi:10.5194/acp-19-9371-2019

Iceberg calving has a strong impact on the internal stress field of marine terminating glaciers and is, therefore, an important indicator for dynamic glacier changes like discharge, acceleration, thinning and retreat. An accurate parameterization of iceberg calving is essential for constraining the glacial evolution and considerably improves simulation results when projecting future sea level contributions. Consequently, temporally and spatially comprehensive datasets of calving front locations are crucial for a better understanding and modelling of marine terminating glaciers. The increasing availability and quality of remote sensing imagery enable us to realize a continuous and accurate mapping of calving front locations. However, the dramatic increase in data volume also accentuates the necessity for automated and scalable delineation strategies.
Due to advances in the field of machine learning, deep artificial neural networks (ANN) are becoming the model of choice for solving complex image processing tasks. Recent studies have already explored the application of these tools for glacier front delineation with very promising results. Rather than simply adding to these studies, we assess the importance of potential input data layers. In particular, we focus on optical Landsat imagery exploiting the full range of multi-spectral capabilities, a statistical textural feature analysis and external topography model data. We estimate their effects on prediction performance through a dropped-variable approach. To do this, we utilize high performance computing systems and re-train our ANN model explicitly removing certain input features. The associated reference dataset comprises more than 1000 satellite images over 23 of the most important Greenlandic outlet glaciers from 2013 to 2021. Resulting feature importances emphasize both the potential in integrating additional input information as well as the significance of their thoughtful selection. We advocate utilizing multi-spectral features, as their integration results in more accurate predictions compared to conventional single-band inputs. This is especially pronounced for challenging ice-mélange, illumination and calving conditions. In contrast, the application of both textural and topographic inputs cannot be recommended without reservation. Their application results in model overfitting which is indicated by a lower accuracy on the validation dataset. The results presented in this contribution reinforce existing efforts for ANN based calving front mapping but also lay the foundation for further applications and developments.

Mapping forest structure at global scale is an important component in understanding the Earth’s carbon cycle. Several new space missions have been developed to support this goal by measuring forest structure predictive of biomass and carbon stock. Furthermore, forest structure characterizes habitats and is thus key for biodiversity conservation. NASA’s Global Ecosystem Dynamics Investigation is one of these missions, which is the first space-based LIDAR designed to measure forest structure (Dubayah et al., 2020). Despite atmospheric noise, the on-orbit full waveforms measured by GEDI are predictive of canopy top height (Lang et al., 2022). Ultimately, these sparse waveforms and derived canopy height metrics will be used to produce global biomass products with 1-km resolution (Dubayah et al., 2020).

Nevertheless, there is a need for high spatial and temporal resolution maps to make informed localized decisions and to improve carbon emission estimates caused by deforestation. Here we present our probabilistic deep learning approach to estimate wall-to-wall canopy height maps from ESA’s optical Sentinel-2 images with a 10 m ground sampling distance. A deep ensemble of fully convolutional neural networks is trained to regress canopy top height using sparse GEDI reference data (Lang et al., 2022). Not only does this approach extend our previous work (Lang et al., 2019, 2021, Becker et al. 2021) from country-level modelling to a global scale, it also yields the predictive uncertainty of the final canopy height estimates. In other words, the model estimates the variance of its predictions indicating in which cases the predictions are less trustworthy.

To enable such a globally trained model to adjust for regional conditions, the geographical coordinates are used as additional inputs to the Sentinel-2 bands. Furthermore, canopy height follows a long-tail distribution, i.e. tall trees are very rare. Thus, a new balancing strategy is developed to reduce the underestimation of tall canopies while preserving the calibration of the predictive uncertainty estimates.

The model performance is evaluated globally on held-out GEDI reference data from randomly selected Sentinel-2 tiles, corresponding to 100 km x 100 km regions. In addition, the resulting maps are compared to dense canopy top height maps (RH98) derived from NASA’s LVIS airborne LIDAR campaigns (AfriSAR, ABoVE/GEDI). On the held-out data the model achieves an RMSE of 5.0 m and a ME of 0.5 m, which indicates a slight overestimation w.r.t. GEDI reference heights. The final, dense predictions are in good agreement with the LVIS derived RH98 and yield an RMSE of 8.8 m and a ME of 0.2 m. Both the usage of geo-coordinates and the balancing strategy reduce the saturation of high canopies. Furthermore, the predictive uncertainty estimates are empirically well calibrated, i.e. the predictive variances correspond to the expected squared errors.

To conclude, the developed methodology makes it possible to produce high-resolution canopy height maps from Sentinel-2 at global scale. How such a model, trained within the GEDI coverage between 51.6° North and South, generalizes to regions north of 51.6° latitude remains to be evaluated with additional reference data.


References:

Dubayah, R., Blair, J. B., Goetz, S., Fatoyinbo, L., Hansen, M., Healey, S., ... & Silva, C. (2020). The Global Ecosystem Dynamics Investigation: High-resolution laser ranging of the Earth’s forests and topography. Science of remote sensing, 1, 100002.

Lang, N., Kalischek, N., Armston, J., Schindler, K., Dubayah, R., & Wegner, J. D. (2022). Global canopy height regression and uncertainty estimation from GEDI LIDAR waveforms with deep ensembles. Remote Sensing of Environment, 268, 112760.

Lang, N., Schindler, K., & Wegner, J. D. (2019). Country-wide high-resolution vegetation height mapping with Sentinel-2. Remote Sensing of Environment, 233, 111347.

Lang, N., Schindler, K., & Wegner, J. D. (2021). High carbon stock mapping at large scale with optical satellite imagery and spaceborne LIDAR. arXiv preprint arXiv:2107.07431.

Becker, A., Russo, S., Puliti, S., Lang, N., Schindler, K., Wegner, J.D., (2021). Country-wide Retrieval of Forest Structure From Optical and SAR Satellite Imagery With Bayesian Deep Learning. Under review

In recent years, numerous deep learning techniques have been proposed to tackle semantic segmentation of aerial and satellite images, trusting leaderboards of main scientific contests and representing today the state-of-the-art.

The encoder-decoder architecture has been widely used for semantic segmentation. Indeed the most popular frameworks for semantic segmentation rely on such an encoder-decoder architecture, e.g. U-Net or Segnet. Such frameworks have been widely used for the semantic segmentation of optical images with very high accuracies.


Nevertheless, despite their promising results, these state-of-the-art techniques are still unable to provide results with the level of accuracy sought in real applications, i.e. in operational settings. They most often perform tasks by learning on examples without having prior knowledge about the tasks. Millions of parameters have to be learned thanks to an optimization process, usually with stochastic gradient descent. Convolutional neural networks have already surpassed human accuracy in many vision tasks. Due to the capacity of convolutional neural networks to fit on a wide diversity of non-linear data points, they require a large amount of training data. Furthermore, neural networks are in general prone to overfitting on small datasets. The model tends to fit well to the training data, but is not accurate for new data. This often makes the neural networks incapable of correctly assessing the uncertainty in the training data and hence leads to overly confident decisions. In order to avoid over-fitting, several regularization techniques have been proposed such as early stopping, weight decay or L1 and L2 regularizations. Currently, the most popular and empirically effective technique to reduce over-fitting is dropout.

Thus, it appears mandatory to qualify these segmentation results and to be able to estimate the uncertainty brought by a deep network. In this work, we address uncertainty estimation in semantic segmentation. Bayesian learning for CNN has been proposed recently and is based on Bayes by Backprop. It produces results similar to traditional deep learning methods, along with uncertainty metrics. In traditional deep learning, models are conditioned on thousand (sometimes millions) of weights w that are learned during training. Once learned, the weights are fixed for further inference. In Bayesian deep learning, models are also conditioned on weights. However, we suppose that each weight follows an unknown distribution. This unknown distribution can be approached by a user-defined variational distribution q(w|theta). Generally this distribution q is a normal distribution and theta denotes the two parameters of the normal distribution, i.e. the mean mu and the standard deviation sigma. However, one can choose any variational distribution for the weights. Hence, unlike traditional networks, the weights of a Bayesian network are not fixed, but conditioned on the variational distribution which parameters are fixed after the learning phase. Thus, the weights can have a wider range of values, allowing the model to learn the data distribution more accurately. Monte Carlo Dropout is equivalent to Bayesian deep learning with its advantages and drawbacks. Its main advantage is that it can be performed by using traditional deep learning optimisation methods (e.g. it is not needed to add the Kullback-Leibler divergence in the cost function). The only condition is to have a learning layer (i.e. convolution layer or dense layer) followed by a dropout layer active in both training and prediction phases. The main drawback is the variational distribution; the user is not able to set the variational distribution. Thus, each weight can only have two values; 0 or a specific value learned during training. Although it seems limited, it is sufficient to learn more accurately the data distribution than a traditional network. In order to have relevant results, several predictions need to be performed in order to explore a sufficient number of values for the weights.

To validate the proposed approach, we consider four different datasets representing various urban scenes. While the first three are public aerial datasets aiming to ease research reproducibility, the last one is a satellite dataset allowing us to demonstrate the behavior of our method on spaceborne imagery as well. The semantic segmentation tasks covers binary classification (building/background) and multiclass classification.

Once trained, a Bayesian model produces different predictions for the same input data since its weights are sampled from a distribution. Therefore, several predictions need to be performed. At each iteration, the model will return a pixel-wise probability. The final semantic segmentation map is computed through a majority vote from all these predictions. One can then derive confusion matrices and usual classification/segmentation quality metrics (precision, recall, accuracy, f-score, intersection over union (IoU) and kappa coefficient kappa). The Bayesian model can also provide uncertainty metrics; two types of uncertainty measures are usually investigated. The Epistemic uncertainty, also known as model uncertainty, represents what the model does not know due to insufficient training data. The Aleatoric uncertainty is due to noisy measurements in the data, and can be explained away with increased sensor precision. These two uncertainties combined form the predictive uncertainty of the network. In this work, we derive two metrics, namely the entropy of the predictive distribution (also known as predictive entropy) and the mutual information between the predictive distribution and the posterior over network weights. These metrics are very interesting since mutual information captures epistemic (or model) uncertainty whereas predictive entropy captures predictive uncertainty which combines both epistemic and aleatoric uncertainties.

Built on the most widespread U-Net architecture, our model achieves semantic segmentation with high accuracy on several state-of-the-art datasets, with with accuracy ranges between 91% and 93%. More importantly, uncertainty maps are also derived from our model. While they allow to perform a sounder qualitative evaluation of the segmentation results, they also appear as a valuable information to improve the reference databases. Furthermore, we showed that our model is very robust to noise, especially when dealing with label noise.

This work has been published in Remote Sensing (https://doi.org/10.3390/rs13193836)

In recent years, deep learning improved the way remote sensing data is processed. The classification of hyperspectral data is no exception. 2D or 3D convolutional neural networks have outperformed classical algorithms on hyperspectral image classification in many cases. However, geological hyperspectral image classification includes several challenges, often including spatially more complex objects than found in other disciplines of hyperspectral imaging that have more spatially similar objects (e.g., as in industrial applications, aerial urban- or farming land cover types). In geological hyperspectral image classification, classical algorithms that focus on the spectral domain still often show higher accuracy, more sensible results, or flexibility due to spatial information independence. DeepGeoMap is inspired by classical machine learning algorithms that focus on the spectral domain like the binary feature fitting- (BFF) and the EnGeoMap algorithm. It is a spectrally focused, spatial information independent, deep multi-layer convolutional neural network for hyperspectral geological data classification. More specifically, the architecture of DeepGeoMap uses a sequential series of different 1D convolutional neural networks layers and fully connected dense layers and utilizes rectified linear unit and softmax activation, 1D max and 1D global average pooling layers, additional dropout to prevent overfitting, and a categorical cross-entropy loss function with Adam gradient descent optimization. DeepGeoMap was realized using Python 3.7 and the machine and deep learning interface TensorFlow with graphical processing unit (GPU) acceleration. This 1D spectrally focused architecture allows DeepGeoMap models to be trained with hyperspectral laboratory image data of geochemically validated samples (e.g., ground truth samples for aerial or mine face images) and then use this laboratory trained model to classify other or larger scenes, similar to classical algorithms that use a spectral library of validated samples for image classification. The classification capabilities of DeepGeoMap have been tested using geochemically validated geological hyperspectral image data sets. The Presentation will include a showcase of how a copper ore laboratory data set was used to train a DeepGeoMap model for the classification and analysis of a larger mine face scene within the Republic of Cyprus, where the samples originated from. DeepGeoMap can achieve higher accuracies and outperform classical algorithms and other neural networks in geological hyperspectral image classification test cases. The spectral focus of DeepGeoMap likely to be the most considerable advantage compared to spectral-spatial classifiers like 2D or 3D neural networks. This enables DeepGeoMap models to train data independently of different spatial entities, shapes, and/or resolutions.

Introduction

This presentation describes the data, method, and results of using SAR data and Deep Learning semantic segmentation (pixel-wise classification) for automated sea ice monitoring. The project was performed at MDA, funded by the Canadian Space Agency, and in collaboration with the Canadian Ice Service (CIS). The goal was to investigate how Deep Learning algorithms could be used to automate and improve SAR-based mapping of sea ice and hence provide more powerful tools for monitoring the impact of climate change on Arctic maritime environments.

At Canadian Ice Service (CIS), image analysts form ice charts (ice type, ice concentration) manually by examination of SAR images, and using their contextual knowledge. However, since this process is time-consuming, the ice charts are often limited to shipping routes and near certain communities. A more extensive mapping, enabled by Deep Learning semantic segmentation, would benefit more communities and allow input to climate models. To facilitate this investigation, the archive of RADARSAT-2 imagery over sea ice, together with the corresponding SIGRID ice charts derived by CIS ice analysts, was used to train Deep Learning models to map sea ice.

As a semantic segmentation problem, the mapping of sea ice is challenging because of the large spatial scale of context and features that influence the classification of a pixel. To overcome this problem, an approach to semantic segmentation using multiple spatial scales was investigated.

Data Gathering and Preparation

CIS uses wide-swath dual-pol (HH-HV) ScanSAR data for ice chart construction, whose 500 km swath provides a large spatial scale for ice features and context. Also the 50 m pixel spacing provides a spatial detail that can be important near inlets and communities. A large number of SAR images and corresponding ice chart data was obtained for the project, and prepared for input to the Deep Learning algorithm. This pre-processing step made use of MDA's Deep Learning pipeline, which provided the following capabilities:
- Image Database to index imagery and metadata
- Image Labelling tool to work with large, geospatial data: conversion and geometric transformation of ice chart data
- Exploitation Ready Product tool for pre-processing SAR imagery: conversion to gamma-zero, dealing with blackfill, geometric transformation
- Dataset Creation tool to create image chips and rasterized label chips for input to Deep Learning algorithm

The project used data from four regions in the Canadian arctic, containing a variety of ice types:
- Middle Arctic Waterways (MID) - 2017 Jul to Oct
- Newfoundland (NFLD) - 2016 Dec to 2017 Jun
- Western Arctic (WA)- 2017 Jun to Nov
- Foxe Basin (FOXE) - 2017 May to 2018 Mar
There were about 30 to 40 ScanSAR image frames for each region.

A CIS ice chart data contains much information about the concentration and the type of the sea ice, which needs to be converted into labels for input to the Deep Learning algorithm. For purposes of this project, we first considered the classification into ice or water, where ice was defined as an ice concentration above 20%. This ice-water classification, at the spatial resolution of the SAR image, provides a detailed description of the ice edge. We also considered the estimation of ice concentration, by classifying the ice into categories corresponding to 20% steps of ice concentration.

The image and ice label data was broken into image chips for input to the Deep Learning model, where the chip size was 512 by 512 pixels, or about 25 km by 25 km, and the chips overlapped by 100 pixels. The image chips are 2-channel chips for the HH and HV polarization. There were between about 8,500 to 21,000 chips per region. The data chips were organized according to image acquisition. For the investigation of the Deep Learning model, the data was split into separate training, validation, and test data sets. The data split was by image, so that all chips from an image were assigned to either training, validation or test, in order to avoid validation on chips that may be similar to training data. The training data was also used to compute image mean and standard deviation, for use in normalizing the input to the Deep Learning algorithm.

Deep Learning Algorithm

The Deep Learning model was implemented using the TensorFlow Estimator framework. The ingest to the model was provided by a generator that read image and label chips, normalized image data, converted label data to the appropriate values for the classification, and computed a sample weight array. The sample weight array is set to zero over land, around the edges, and at NaN image values, and is used to compute the weighted loss function for training and the weighted metrics for evaluation.

The Deep Learning model was based on the DeepLab model for semantic segmentation, which uses dilated convolutions to provide large convolution kernels for context, without decimation of the image. The DeepLab model is also built using ResNet blocks which improves training for large networks.

For an input image chip, the output of the model is an array of predicted label values which has the same size as the input chip. During training, the predicted labels and the true labels are weighted by the sample weight array, and then used to compute the Cross Entropy loss function for optimization of the model weights. During validation, the predicted labels and the true labels are weighted and used to compute the accuracy and the mean intersection-over-union, which is a common metric for semantic segmentation.

In addition to the original Deep Learning model for semantic segmentation, a new approach was developed to apply Deep Learning at multiple spatial scales. This was done to overcome the problem of sea ice features that extended beyond the field-of-view of the model that could be provided by a single chip. In this approach the Deep Learning model takes multiple inputs, one of which is the original image chip, the other is the result of classification using down-sampled data, which provides a larger context.

Results

The performance of the DeepLab model for ice-water classification was assessed. First, the data for each of the four regions was used separately to train and evaluate different models. The accuracy and mean intersection-over-union (mIOU) are:
Mid-Arctic: accuracy = 0.9175, mIOU = 0.8322
Newfoundland: accuracy = 0.9582, mIOU = 0.9194
Western Arctic: accuracy = 0.9675, mIOU = 0.8869
Foxe Basin: accuracy = 0.9334, mIOU = 0.8611
Then, the data was combined to train one combined model, and the combined model was used to initialize the training for each region, in a process called fine-tuning. The fine-tuning provided the best result, which compared to separate training improved accuracy on average by about 1%, and improved mean intersection-over-union by about 2%. The figure shows an example of the ice-water classification in the Western Arctic region, using the fine-tuning strategy. The figure shows the HH and HV images, the true labels, and the predictions.

The performance of the DeepLab model for ice concentration classification was assessed. Each of the 5 classes represented a different ice concentration: 0-20% (water), 20-40%, 40-60%, 60-80%, and 80-100% (ice). Since the most of the image pixels were ice or water, with relatively few pixels at intermediate concentrations, there was less data for training these intermediate concentration classes. The effect of this limitation can be seen in the results. Whereas the mIOU values for the ice and water classes were above 0.75 for most of the regions, the mIOU values for the intermediate concentrations were typically below 0.2.

Finally, the approach of using multiple inputs at different spatial scales for semantic segmentation was investigated. This was done for the case of ice-water classification, using separate training for each region. The steps in this approach are:
- down-sample the original image
- train a model using the down-sampled data
- form predictions on the down-sampled data
- for each original resolution chip, extract the prediction on the down-sampled data corresponding to the same area
- train a model using both the original resolution chip, and the down-sampled prediction

The multiscale approach was investigated using 4-times down-sampling, using the Mid-Arctic and Newfoundland regions. The results on ice-water classification are:
Mid-Artic: accuracy = 0.9320, mIOU = 0.8613
Newfoundland: accuracy = 0.9602, mIOU = 0.9243

In particular, there were certain image frames for which the original performance was poor due to a lack of context within a single chip. This was especially true for some of the Newfoundland data. For these difficult images, the improvement in accuracy and mIOU using the multi-scale approach was over 7%.

Conclusion

The performance of the algorithm is very promising, indicating that Deep Learning semantic segmentation has a lot of potential to aid in the automation and improvement of sea ice mapping.

In times of rising world population, increasing use of agricultural products as energy sources, and climate change, the area-wide monitoring of agricultural land is of considerable economic, ecological, and political significance. Crop type information is a crucial requirement for yield forecasts, agricultural water balance models, remote sensing based derivation of biophysical parameters, and precision farming. To allow for long enough forecast intervals that are meaningful for agricultural management purposes, knowledge about types of crops is needed as early as possible, i.e. several months before harvest. Thus, such early-season crop-type information is relevant for a variety of user groups such as public institutions (subsidy control, statistics) or private actors (farmers, agropharma companies, dependent industries).

The identification of crop types has been a long research topic in remote sensing, starting from mono-temporal Landsat scenes in the 1980s to multi-sensor satellite time series data nowadays. However, most often crop types are identified in a late cultivation phase or retrospectively after harvest. Existing products are mainly static and not available in a timely manner and therefore cannot be included in the decision-making and control processes of the users during the cultivation phase.

We currently develop a web-based service for dynamic intra-season crop type classification using multi-sensor satellite time series data and machine learning. We make use of the dense time series the Copernicus Sentinel satellite fleet offers and combine optical (Sentinel-2), and SAR (Sentinel-1) data providing detailed information about the temporal development of the phenological state of the crop growing phases. This synergetic use of optic and radar sensors allows a multi-modal characterization of crops over time using passive optical reflectance spectra and SAR-based derivatives (i.e. backscatter intensities and structural parameters derived by polarimetric decompositions). The automatic data processing pipeline of data retrieval, data pre-processing, and data preparation as a prerequisite for applying machine learning algorithms is based on open-source tools using SNAP and python libraries as main functionalities.

The developed AI-based model uses the multi-modal remote sensing time series data stream to predict crop types early in their growing season. This model is based on previous work by Garnot et al. 2020, who leverage the Attention mechanism originally introduced in the famous Transformer architecture in order to better exploit the information about crop type included in the change of appearance in satellite images over time. The original model focuses on prediction of crop type based on Sentinel-2 acquisitions from within a single Sentinel-2 tile, which leads to very similar acquisition time points for all parcels in the dataset. This is not the case when applying the model to larger regions or including other data sources, such as Sentinel-1 (polarimetric and backscatter), which have vastly different acquisition time points.

By implementing a modified form of positional encoding, we are able to train and predict on regions and data sources with differing acquisition time points as we provide implicit information about the acquisition time point directly to the model. This means we don’t need any temporal data preprocessing (e. g. weekly/monthly averages) and allows us to seamlessly fuse data from different sources (Sentinel-1 and Sentinel-2), leading to good prediction performance also in periods where there is no Sentinel-2 data available due to cloud occlusion.

In order to improve the generalisation abilities of our model across regions and different years, we also study the effect of fusing satellite data with geolocalised temperature and precipitation measurements to account for the dependence of growth periods on these two parameters.

We will present insights on the developed dynamic crop type classification service based on the use case for the federal states Mecklenburg-Vorpommern and Brandenburg in Germany. For both states, official reference information from the municipalities about the cultivation information for approx. 200,000 fields are used for training and testing the algorithm. Predicted crop types are winter wheat, winter barley, winter rye, maize, potatoes, rapeseed, and sugar beet. We will show model performances in different cultivation stages (from early season to late season) and with different remote sensing data streams by using Sentinel-1 or Sentinel-2 data separately or in conjunction. Moreover, the transferability of the approach will be evaluated by applying a trained model of one year to other years not included in the training phase.

Worldwide economic development and populations growth has led to unprecedented urban area change in the 21st century. Many changes to urban areas occur in a short period of time, raising questions as to how these changes impact populations and the environment. As satellite data become available at higher spatial (3-10 m) and temporal (1-3 days) resolution, new opportunities arise to monitor changes in urban areas. In this study, we aim to detect and map changes in urban areas associated with anthropogenic processes (for example, constructions) by combining high-resolution Sentinel-2 data with deep learning techniques. We used the Onera Satellite Change Detection (OSCD) dataset containing Sentinel-2 images pairs with changes labeled across 24 locations. We advanced the OSCD by implementing state-of-art algorithms for atmospheric correction and co-registration of Sentinel-2 images, as well as improve the deep learning model by introducing a modified loss function. A new test area, named Washington D.C., was used to demonstrate the model’s robustness. We show that the performance of the model varies significantly depending on the location: F1 score ranges from 3.37% in West Saclay (France) to 74.16% in Rio (Brazil).

The developed model was applied to mapping and estimating area of changes in several metropolitan areas, including Washington D.C. and Baltimore in the US in 2018-2019 and in Kyiv, Ukraine, 2015-2020. A sample-based approach was adopted for estimating accuracies and areas of change. Stratified random sampling was employed, where strata were derived from change detection maps. Since in our cases, the area of no change would account for >99% of the total area, the corresponding weight would influence the derived uncertainties of area estimates (for example, see Eq. (10) in Olofsson et al. (2014))—the larger the stratum weight, the more considerable uncertainties of estimated areas of change. Therefore, a spatial buffer of 20 pixels (at 10 m) was introduced to include areas of no change around areas of change. The main goal of introducing the buffer was to mitigate the effects of omission errors, which would lead to large uncertainties in change area assessment (Olofsson et al., 2020). Overall, 500 samples were used for each location (DC, Baltimore and Kyiv), with 100 samples allocated for the change stratum, 100 samples for the buffer stratum, and the rest 300 samples allocated for the no-change stratum. In terms of response design, a 10-m pixel was selected as an elementary sampling unit. The reference data source was corresponding imagery available in Google Earth.

We estimated that in just one year, 2018-2019, almost 1% of the total urban area in DC and Baltimore underwent change: 10.9±4.3 km2 (0.85% of the total area in DC) and 10.8±2.2 km2 (0.92% in Baltimore). Among detected changes, active constructions (those that can be seen in the 2019 imagery) accounted for 78% and 86% in DC and Baltimore, respectively, while the rest represented the completed constructions. Commercial buildings accounted for 52% and 46%, and residential buildings accounted for 27% and 21%. Worth noting that 8-9% of detected changes in DC and Baltimore occurred due to the construction of new schools or renovation of existing ones. This high number can result from the growing density population and the number of residential properties being built (as shown in this study), as such overcrowded school buildings requiring renovations. Another type of change identified was the construction of parking lots next to commercial buildings, roads, and hospitals.

In Kyiv, the area of change was estimated at 17.0±2.8 km2 between 2015 and 2020 that constituted 2.1% of the total area. Active constructions accounted for 38%, while the rest were visible finished constructions. Constructions with primary residential land use zoning accounted for 40%, while commercial buildings accounted for 15%.

This study highlights the importance of the overall framework for urban area change detection: from building models using benchmark datasets to actual mapping and change area estimation through sample-based approach, which provides unbiased estimates of areas.

Being able to reconstruct 3D building models with a high Level of Details (LoD) — as described by the CityGML standard from the Open Geospatial Consortium — from optical satellite images is needed for applications such as urban growth monitoring, smart cities, autonomous transportation or natural disaster monitoring. More and more Very High Resolution (VHR) data having a world wide coverage are available, and new VHR missions for Earth Observation are regularly launched as their cost decreases. One key strength of satellite imagery over aerial or LiDAR data is its high revisit frequency allowing to better track urban changes and update their 3D representations. However, extracting and rendering buildings with finer precision, notably regarding their contours, in an automatic fashion still remain a challenge for lower spatial resolution (1 m to 50 cm), as well as minimizing manual post-processing and enabling large scale application.

This work presents an end-to-end LoD1 automatic building 3D reconstruction pipeline from multi view satellite imagery. The proposed pipeline is composed of different steps including a multi-view stereo processing with the French National Centre for Space Studies (CNES) tool named CARS to extract the Digital Surface Model (DSM) followed by the generation of the Digital Terrain Model (DTM) from the DSM by an image implementation of the Drap Cloth algorithm. A building footprint extraction step is carried out through semantic segmentation with a deep learning approach. These footprints are vectorized, regularized with a data-driven approach and geocoded so to form the LoD0 reconstruction model. Eventually, the regularized shapes are extruded from the ground to the gutter height to lead to the LoD1 reconstruction model. Attention is drawn to the building contour restitution quality.

Our proposed pipeline is evaluated on Pleiades multi-view satellite images (GSD of 50cm). Deep learning networks are trained on Toulouse, France, and the pipeline is evaluated on several other French urban areas which exhibit spectral, structural and altimetric variations (such as Paris). Building footprints groundtruth are extracted from the OpenStreet Map project. State of the art results are achieved for building footprints extraction. A good generalization capacity is highlighted without need of a new training for machine learning models.

Human activity is the leading cause of wildfires, however, lightning can also contribute significantly. Lightning ignited fires are unpredictable and identifying the relationship between lightning and its causes for ignition is useful for fire control and prevention services, who currently assess this danger mainly based on forecasts of local lightning activity. In general, there is a relationship between fire ignition and fuel availability and dryness, but the environmental conditions needed for a lightning ignition was unknown. This research looked at developing a global artificial intelligence (also known as machine learning) model which would be triggered by cloud to ground lightning activity and then examine the current weather and ground conditions to identify lightning flashes which could potentially cause fires, adding valuable information to already existing regional early warning systems.

In this research we created three different machine learning models called classifiers to identify where lightning flashes could be a hazard. This model type assigns a label to a series of information given and for these research models, it assigned one of two: lighting-ignited fire hazard or no hazard. All models were developed on the decision tree concept, a basic model, where you follow a series of questions, answering true or false until you reach a decision. The models developed were a singular decision tree, Random Forest and Adaboost, with the latter two methods using many different decision trees to create an answer. The models were built on environmental data gathered from assumed past lightning ignited fires events identified by combining active fire information with a cloud to ground lighting forecast product.

Original test data showed promising results of around 78% accuracy for the multiple decision tree methods, leading to an independent verification 145 lightning ignited fires in Western Australia in 2016. This highlighted that in a minimum of 71% of the cases the models correctly predicted the occurrence of lightning caused fire. Due to the success of the models, further research is planned with the current models to be used in an operational context to enhance information connected to fire management.

Societal challenges like climate change and adaptation, digitalisation, mobility, renewable energy, sustainability and civil security are the main topics, for which the German Earth Observation Programme is striving to deliver contributions and solutions. As the largest contributor to the European Earth Observation Programmes at ESA, Eumetsat and EC, Germany is shaping together with our international partners a user-driven, well-balanced, technological outstanding and long-term EO landscape for Earth science, operational services and downstream business. The German National EO Programme complements the European activities in terms of data exploitation, service development, technology preparation and mission development and operations. The TerraSAR-X and the TanDEM-X missions as the closest-ever in-orbit formation delivered a digital elevation model of all land masses enabling Earth System Science based on a homogeneous topographic data base. The German-US GRACE FO mission already demonstrated its capability to continue and enhance the successful GRACE mission. DESIS is delivering hyperspectral data from the ISS. With the hyperspectral EnMAP mission and the German-French MERLIN mission further elements of a German mission portfolio, complementary to ESA’s Earth Explorers, to the Copernicus Sentinels and to the EUMETSAT missions will be placed in orbit in the next years. In addition, with the in-kind contribution of the MetImage instrument to EUMETSAT’s post-EPS satellite system this operational satellite series will be able to provide indispensable data for Earth Observation applications. At the same time, new trends in EO like New Space, online platforms and cloud processing, AI and HAPS are offering new potentials for service opportunities, which are explored in our programme.
With the long-term data and service continuity as one of the major user requirements the strategy of Germany for its next generation of high-resolution radar, gravity, Lidar, optical, infrared and hyperspectral space activities is of utmost importance. These plans and the role of these missions in the future global Earth Observation System will be introduced.

The “Advanced Optical Satellite” (ALOS-3, nick-named “DAICHI-3”) is the next high-resolution optical mission as a successor of the optical mission by the Advanced Land Observing Satellite (ALOS, “DAICHI”) in Japan Aerospace Exploration Agency (JAXA). ALOS-3 is now under testing the flight model and the launch preparations will be completed soon. ALOS-3 is scheduled to be launched by March 2022.
The major mission objectives are (1) to contribute safe and secure social including provisions for natural disasters, and (2) to create and update geospatial information in land and coastal areas. The “WIde-Swath and High-resolution optical imager” (WISH, as a tentative name) will be mounted on ALOS-3, which consists of 0.8 m resolution of the panchromatic band and 3.2 m resolution of multispectral six bands with 70 km of observation swath width. This paper describes overviews of ALOS-3’s mission and products, and the updated calibration and validation plan of WISH, which will be conducted after the launch.
Regarding two major mission objectives of ALOS-3, JAXA expects to utilize the following various applications and outcomes by ALOS-3.
1. Safe and secure social including provision for natural disasters
For responding to natural disasters not only in Japan but also worldwide, disaster-related information e.g., damaged area and volume estimations, damage assessment associated with rescue activities will be provided as soon as after happening the event. To accommodate this requirement, ALOS-3 had requested its observation agility. For analyzing the acquired data, this is a change detection between before and after the event, therefore an observation repeatedly and the archiving data is important before the event as well. This is also able to contribute to maintaining and updating the hazard maps in the prevention phase. In the emergency phase, multi-satellite responses by ALOS-3 and ALOS-2 if it is still in operating, and ALOS-4 as the next Synthetic Aperture Radar (SAR) mission will be considered for the quick response and valuable information extractions.
2. Geo-spatial information in land and coastal areas
The Geospatial Information Authority of Japan (GSI) is responsible for official national topographic map generation and update with 1/25,000 scales. To contribute to this requirement, at least 5 m geometric accuracy must be guaranteed. To identify surface textures, land-use and land-cover (LULC) and their changes to update the map as well, image readability and quality are also important. In addition, terrain height estimation i.e., digital elevation model (DEM) or digital surface model (DSM) is important to create contour lines in the map. This function may also contribute to activities in the natural disaster responses.
The latest information including the results of the launch and the Initial Check Out will be presented in the presentation.

The paper presents the development status of PLATiNO platform with first four payloads and missions.
PLATiNO is a project initiated by the Italian space Agency with the aim of developing a multi-mission and multipurpose flexible and scalable platform developed by the industrial consortium composed by SITAEL, Thales Alenia Space in Italy, Leonardo and Airbus- Italia..
The multi-applicability has been proven through a joint and parallel development of a SAR mission and an optical Thermal Infrared Mission whose design is now completed and the program is undergoing the qualification phase. In addition, ASI has decided to develop additional optical EO missions to accomplish the national Optical Roadmap for the Earth Observation. Therefore, the national mission portfolio, based on PLATiNO platform, spans among a SAR mission (namely PLT-1), a TIR mission (namely PLT-2), a VNIR mission (namely PLT-3) and a Hyperspectral mission (namely PLT-4).
PLATiNO platform is a new all-electric propulsion small platform product in the mini-satellites class with total mass falling in the range of 250-350 kg (S/C launch mass), designed to be compatible with a wide range of applications (multi-applicability feature). The platform design features and technological solutions (i.e. electric propulsion for V-LEO orbits, mini-CMG for agile re-pointing, ISL for formation flying/constellations, high data rate active antenna for EO Data management) are at the state-of-art and strictly linked to the multi-purpose capability. While the platform is a standard product compatible with several payloads (EO – SAR, Optical – TLC – Science), the first four missions are all related to Earth Observation to meet the needs of the Agency in consolidating and expanding its capabilities in investigating and monitoring the changes and their causes occurring onto the Italian territory and the planet in general.
PLATiNO payloads and missions are presented, with focus on each of their specific performances.
The first mission, PLT-1, is a high resolution and compact SAR mission, that will be launched at the begin of 2023. PLT-1 SAR can perform stripmap bi-static and monostatic imaging, as well as Spotlight modes flying in formation with a Cosmo SkyMed satellite.
The second mission, PLT-2, is a hi-res Thermal InfraRed mission that will be launcher at the beginning of 2024 and which is able to observe the Earth both during day and night due to the chosen spectrum which covers the IR between 8 and 12 microns, i.e. in the emissivity region. The orbit will be an SSO with an altitude lower than 400 km.
The third mission, PLT-3, is a Visible and Near InfraRed Mission to be launched in mid 2025, which represents a step ahead in the Italian EO capabilities aimed at observing the Earth at high resolution and wide swath.
The fourth mission, PLT-4, is a Hyper Spectral mission to be launched in mid 2026 aimed at consolidating the Italian Hyperspectral observation capabilities also by the utilization of a small and agile platform.

Vegetation and Evironment New Micro-Satellite (VENµS) is an Earth observation space mission jointly ‎developed, manufactured, and operated by the National Centre for Space Studies (CNES, France) and ‎the Israel Space Agency (ISA). The satellite, launched in August 2017, crosses the equator at around ‎‎10:30 AM Coordinated Universal Time (UTC) through a sun-synchronous orbit at 720 km height with 98° ‎inclination. During its first phase, named VM1, the scientific goal of VENμS was to frequently acquire ‎images on 160 preselected sites with a two-day revisit time, a high spatial resolution of 5 m, and 12 ‎narrow bands, ranging from 424 to 909 nm. This band setting was designed to characterize vegetation ‎status, monitor water quality in coastal and inland waters, and estimate the aerosol optical depth and ‎the water vapor content of the atmosphere. To observe specific sites within its 27-km swath, the ‎satellite can be tilted up to 30 degrees along and across track. Uniquely, the preselected sites are ‎always observed with constant view azimuth and zenith angles. Four spectral bands were carefully set ‎in the red-edge region between the atmospheric absorption areas. Also, exceptionally, there are two ‎identical red bands located at both extremities of the focal plane. The 2.5-sec difference between the ‎first and the last red bands of the pushbroom scanner enables stereoscopic view and retrieves 3-‎dimensional measurements.‎

The presentation strives to demonstrate several applications derived from the VENµS unique ‎characteristics. The frequent revisit time enables creating a dense, high-quality, time series of crops ‎and, thus, accurately depicts different phenological stages (e.g., sowing, germination, vegetative, ‎mortality). It also enables the detection of near-real-time changes of constituents in water bodies, ‎glacier flow, and more. The red-edge inflection point (REIP) was proven to be a better index than ‎NDVI for field crops studies, including predicting leaf area index and chlorophyll and nitrogen contents. ‎The spectroscopic capability allows retrieving a digital surface model (DSM). DSM provides the altitude ‎of the natural terrain and features on the Earth's surface, such as trees, buildings, etc. Stereoscopy ‎also enables cloud classification based on their heights and enhances cloud-mask algorithms. The ‎multi-angular view capability is used for improving vegetation (and other ground features) monitoring ‎and modeling. ‎

VENµS is also used to prepare the next optical satellite missions to help determine the optimal revisit ‎time. For instance, it has been shown that the two-day revisit time enables the production of bi-‎monthly syntheses with a limited number of residual clouds. Additionally, VENµS is used to optimize ‎the revisit of a companion mission to Sentinel-2, named Sentinel-HR, that would bring metric or bi-‎metric resolution with a reduced revisit (~20 days). Consequently, VENµS is useful to benchmark the ‎spatio-temporal fusion methods for merging the benefit of the Sentinel-2 and Sentinel-HR missions ‎and getting metric resolution images with a revisit of two days. These advantages will be enhanced ‎during the new phase of the mission, termed VM5, for which the orbit of VENµS will be at 560 km ‎altitude, with a new unique combination of features. i.e., a one-day revisit time and 4 m spatial ‎resolution. VM5 is planned to last two years, starting from January 2022.‎

Imaging spectroscopy in the VNIR/SWIR spectral range has demonstrated strong potential for the characterization of chemical and physical properties of Earth surface materials and processes. Earth observation applications based on imaging spectroscopy include the characterization of vegetation properties in natural and managed systems, top soil properties, water properties in coastal areas and inland water, urban land cover, industrial waste and air pollution or soil contamination. Following the Hyperion mission, space missions recently became operational (PRISMA), others will be launched soon (EnMAP), and global missions are under study (CHIME, SBG). Most of them have a ground sampling distance (GSD) of 30 m, a large swath and can therefore cover large areas on Earth to characterize different terrestrial and oceanic ecosystems with a revisit period ranging from 4 to 16 days. Such a spatial resolution is a limiting factor for an accurate discrimination of heterogeneous areas and the characterization of specific ecosystems, because it induces a large number of mixed pixels. The BIODIVERSITY (ex-HYPXIM) mission aims at complementing these space missions with a unique combination of characteristics including a GSD of 10 m, a revisit time of up to 5 days and a spectral range from 0.4 to 2.4 µm. It will thus provide answers to several scientific issues (e.g., Biodiversity monitoring, shallow water biodiversity monitoring, soil contamination monitoring...), which motivated the conception of the instrument.
To support BIODIVERSITY, the French scientific community focused on identifying the requirements for spectral resolution, radiometric resolution and absolute calibration, evaluated based on a set of applications covering aforementioned topics. For each topic, the illustration and the performance on the estimated variables are presented and the best configurations is deduced. These applications include:
• The characterization of vegetation traits in tree-level species assemblages; these traits are associated with the resilience of terrestrial ecosystems, anthropogenic influences, and ecosystem biodiversity in terms of species composition and assemblages. Illustrations will be given on the estimation of Essential Biodiversity Variables (EBV): classification of species of temperate forest and estimation of pigments (Chlab, carotenoids), leaf water content and Leaf Mass Area (LMA) of Mediterranean forest.
• The improved knowledge on biodiversity and bathymetry in shallow water for coastal areas and inland waters. Results will focus on the estimation of shallow water biodiversity and bathymetry.
• The characterization of top soil properties to assess soil pollution and soil quality at fine spatial resolution, providing information on the influence of soil management practices on environmental processes such as soil carbon sequestration, infiltration and retention, runoff and soil erosion. This will be illustrated with mineral discrimination and Soil Moisture Content (SMC) estimation.
• Imaging spectroscopy can monitor cities and industrial pollution to evaluate the urban sprawl or the quality of our environment. We show that this GSD will improve our understanding of urban areas and activities of industrial sites to retrieve the urban land cover, or the solid and liquid effluents and the atmospheric productions, aerosols and greenhouse gases, of industrial activities.

Landsat 9 is a partnership between the National Aeronautics and Space Administration (NASA) and the U.S. Geological Survey (USGS) that will continue the Landsat program’s critical role of repeat global observations for monitoring, understanding, and managing Earth’s natural resources. Since 1972, Landsat data have provided a unique resource for those who work in agriculture, geology, forestry, regional planning, education, mapping, and global-change research. Landsat images have also proved invaluable to the International Charter: Space and Major Disasters, supporting emergency response and disaster relief to save lives. With the addition of Landsat 9, the Landsat program’s record of land imaging will be extended to over half a century.

The successful launch of Landsat 9 from Vandenberg Space Force Base, California, USA on September 27, 2021 onboard a United Launch Alliance Atlas V 401 rocket represented a major milestone for a five-decade partnership between NASA and the USGS that continues to set the standard for high-quality Earth observations. During the 100-day commissioning phase, NASA monitored all aspects of the spacecraft as it traveled toward its final orbit height of 705 km (438 mi.) above the Earth. Spacecraft maneuvers and calibration were conducted throughout the commissioning phase to verify that all systems were operating nominally. At approximately 100 days, ownership of the Landsat 9 mission was transferred to the USGS which began the operations phase.

Landsat 9 collects as many as 750 scenes per day, and along with Landsat 8, the two satellites add nearly 1,500 new scenes each day to the USGS Landsat archive for access by the global user community from a free and open cloud-based architecture. Processed into USGS Landsat Collection 2, the Landsat 9 products promise to be more interoperable than ever with other datasets, such as those offered by the Copernicus Sentinel-2 missions. Additionally, the global network of Landsat international ground stations provides contingency operations and make available near real-time Landsat products to serve local and regional user needs.

Ocean surface current observations are essential for monitoring upper-ocean transport of heat, nutrients, pollutants, validation of model forecasts, data assimilation, and marine operations. Existing surface current measurements from in situ drifters and moorings, shore-based radars, and satellite altimetry are either not available on a regular basis, cover only limited areas, and/or are not applicable near the coast. The Doppler shift acquired by satellite Synthetic Aperture Radars (SARs) over the ocean is a measure of the radial total surface motion induced by the near-surface wind, surface waves, and underlying surface currents. Given accurate removal of the sea state contributions, such data can be used to retrieve global surface current radial velocities (RVLs) with a high spatial resolution of 1 km. In this study, we developed an empirical Geophysical Model Function (GMF) for predicting the sea state contribution to the Doppler shift in order to improve the accuracy of the ocean surface current radial velocity retrievals in the coastal zone.

The Sentinel-1 mission is an operational constellation of two C-Band SAR instruments (A/B) launched in 2014/2016. Challenging calibration has prevented their usage for retrieving surface currents. Recently an experimental calibration emerged thanks to two months of telemetry from the gyroscope onboard Sentinel-1B, recalibrated on-ground, yielding promising improvement in the accuracy of the geometric Doppler shift component. Following this development, we generated experimental Sentinel-1B IW RVL products over the Norwegian coastal zone in December 2017 - January 2018 and collocated them with the wind and wave fields from the regional operational MEPS and MyWaveWAM models. We estimated that the Doppler shift from the experimental products has a better accuracy of about 3.8 Hz compared to the 6.8 Hz previously reported for standard products. We further trained the model function (CDOP3SiX) that predicts the sea state Doppler shift as a function of the range-directed wind, wind sea orbital velocity, swell orbital velocity, as well as incidence angle. As such, the CDOP3SiX accounts for a more realistic sea state representation compared to the previously used empirical models (e.g., CDOP) that relied only on access to the wind fields and assumption of fully developed wind sea in absence of swell. We found that the signal from the Norwegian Coastal Current of about 0.5 m/s can be systematically detected in the Sentinel-1 derived ocean surface current radial velocity fields with 1 km spatial resolution. Moreover, the Sentinel-1 derived surface currents also express the presence of meandering structures and boundaries consistent with the satellite-based sea surface temperature field. Comparison with the ocean model also reveals acceptable agreement, especially for the major surface current features.

Despite the study being constrained by only two months of data from the single Sentinel-1B satellite, the results are promising. Reprocessing of the full Sentinel-1 A/B dataset using novel attitude calibration is therefore essential for further improvement of the empirical algorithms and validation. The developed methodology can be applied for the observations from the operational Sentinel-1 mission (sustained operation until 2030) as well as adapted for candidate future satellite missions designed for monitoring of the upper ocean circulation (e.g., SEASTAR and Harmony). However, CDOP3SiX relies on the usage of collocated wind and wave forecasts that introduces corresponding errors and limits its application to the various regions. Therefore, the next generation of empirical models might explore the possibility to retrieve wind and wave fields directly from the SAR observations, thus avoiding the use of numerical model fields that cannot realistically represent the high spatial variability in the wind and wave conditions that are typical for given SAR acquisitions. In turn, a highly valuable dataset of Doppler-based radial velocities would stimulate more advanced studies of the upper ocean dynamics and comparison to numerical ocean model simulations and predictions, and, eventually, assimilation of the SAR-derived surface current retrievals.

The Doppler Centroid (DC) frequency shift recorded over ocean surfaces by Synthetic Aperture Radar (SAR) is a sum of contributions from satellite attitude/antenna and ocean surface motion induced by waves and underlying ocean currents. A precise calibration of the DC is needed in order to predict and subsequently remove contributions from attitude/antenna.
Recently, a novel data calibration technique based on combining gyroscope telemetry data and global Sentinel-1 WV OCN products (OceanData Lab Ltd, 2019), has demonstrated promising capabilities to quantify the Sentinel-1 (S1) attitude and hence provide calibrated estimates of the corresponding DC frequency shift. One year of S1 a and b WV OCN products, orbit data and gyroscope data are combined providing one year of restituted attitude data (AUX_ESTATT). For the same time period, mean DC bias versus elevation angle is computed on a daily basis from S1 IW land acquisitions (AUX_DCBIAS). The AUX_ESTATT and AUX_DCBIAS products are subsequently used to generate global data set of calibrated S1 WV OCN products as well as sub-sets of calibrate S1 IW OCN products from predefined super sites (Norwegian Coast, Agulhas, Mediterranean). In this paper we assess the accuracy and precision of the calibrated DC frequency of S1 WV and IW acquisitions acquired over both land and ocean areas. The DC standard deviation (STD) and bias show significant reduction for both satellites and for all swaths. Assessment of the performance of global WV data shows a STD around 6Hz, while the BIAS is less than 2 Hz. The performance is very similar for both satellites and for both swaths. For IW the STD is similar, but the bias is slightly higher and small DC bias between sub-swaths is sometimes observed.
The remaining errors are mainly due to change in antenna characteristics on a timescale not capture with the procedure used to generate the mean DC bias stored in the AUX_DCBIAS file. Such changes may come from thermo-elastic effects and/or temperature compensations applied to the antenna. This directly affects the IW mode DC, where it is also clearly visible in some scenes. For WV mode it mainly impacts the statistics.
We conclude that S1 WV mode has achieved a performance (i.e. accuracy and precision) within the requirement for climatology mapping of global ocean current features. For IW mode, we have achieved a precision within the requirement, but use of land areas within the scene is still required to achieve the required accuracy over all sub-swaths.

The ocean meso- and submesoscale dynamic processes are one of the gaps in our understanding of the ocean-atmosphere exchange of momentum and energy. Resolving the ocean currents and waves at submesoscale resolution is one of the missing pieces in the dynamic ocean-atmosphere processes puzzle. Harmony, an Earth Explorer 10 candidate currently in phase-A studies, proposes direct instantaneous Doppler frequency shift measurements of the ocean surface currents, surface wind stress, and wave spectra.

Harmony achieves the aforementioned measurements using two spacebourne synthetic-aperture radar (SAR) companion satellites flying together with Sentinel-1. Each of the two companions will form a bistatic SAR with Sentinel-1 as the illuminator, allowing for Doppler measurements of ocean surface velocities with two lines of sight and along-track interferometry (ATI). The novel acquisition geometry poses challenges in the modelling of measurement errors that are not addressed with methods for monostatic SAR systems.

In this study we present a quantitative assessment of the errors involved in the Doppler measurement of ocean surface vectors from bistatic, formation-flying SAR, particularly as it pertains to the Harmony mission. Specifically, we investigate the effects of two types of errors: random measurement noise and systematic errors related to the instrument and the satellite. Moreover, we present algorithms to correct for the systematic errors and evaluate calibration techniques for ocean Doppler measurements.

Measurement noise is typically the better understood type of error out of the two in the field of SAR Doppler measurements. It is modelled as thermal white noise driven by the ocean backscatter, the coherence of the measurement and the number of independent looks. The coherence is a function of the NESZ, the temporal baseline and the ambiguities. Mitigation thus can only be achieved by improving the instrument sensitivity and resolution during the design phase or by adjusting the baseline in the case of ATI. We present the impact of measurement noise on interferometric performance and the trade-offs in instrument design.

Systematic errors on the other hand, arise due to unknown perturbations in the output signal of the instrument driven by the electronics of the receiver and the processing chain, and by uncertainties in the position of the antenna phase centres and the position of the companions in the formation. The perturbations are a result of signal ambiguities and clock errors, while the position uncertainties can be due to attitude errors, structural vibrations and reference-height errors. All systematic errors materialise as an undesired phase offset in the Doppler measurement that needs to be estimated and calibrated for.

Calibration of a SAR instrument over the ocean surface is difficult due to the dynamic nature of the surface. Thus, calibration is in practice done over non-moving targets such as land. For such a technique to work, systematic errors must have a drift that stays as small as possible during acquisitions over the ocean. Understanding the spatial scale of the error variance and the rate at which decorrelation occurs is important in determining effective mitigation techniques for Harmony. Recognising the spatial dependence of the systematic errors, we propose adopting methods from the field of Spatial Statistics to better understand the uncertainties.

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 their variabilities under climate change. Presently, measurements of global ocean surface currents, which are mainly geostrophically derived from satellite altimeter data, are only available to resolve quasi-geostrophic current at large- to meso- scale in the off-equatorial open ocean. Ocean Surface Current multiscale Observation Mission (OSCOM) will launch a satellite equipped with a Doppler Scatterometer to directly measure ocean surface currents with a very high horizontal resolution of 5–10 km and a 3-day global coverage. OSCOM will provide an in-depth picture of non-equilibrium ocean state and air-sea interaction from mesoscale to submesoscale, and helps to construct the fine structure of deep ocean current through a combination with Argo profiling. Those direct measurements and derived dynamic parameters will further provide a novel and improved pathway to data assimilation and coupling of GCMs for ocean prediction and climate change.
OSCAR current and the surface drifters indicate that the non-geostrophic currents in the global ocean account for ~43% of the total current. Especially, the non-geostrophic currents determine the directions of the total currents in the near-equatorial trade winds and mid-latitude westerly winds prevailing regions, where the maximum non-geostrophic speed can reach twice the geostrophic speed and exceed 60% of the total current. The present current reanalysis cannot reveal the non-geostrophic processes in these regions and underestimate the weakening effect of the non-geostrophic process in the strong western boundary currents and the Antarctic Circumpolar Current. The influence of the non-geostrophic processes in the real world will be more significant, and the OSCOM is expected to reveal these processes in the future.

Ocean surface current is an essential ocean and climate variable, and it plays an important role in various scientific research and engineering applications. At present, there is only global geostrophic current derived from spaceborne altimetry, however, no direct global total ocean surface current vector observations from space are available. Geostrophic current is part of contribution to the total ocean surface current, and there are still ageostrophic current processes, however, the contribution of geostrophic current and ageostrophic current to the total ocean surface current is still unclear. Besides, retrieving the total ocean surface current is still challenging, as the total Doppler shift from ocean surface includes the contributions from sensor platform motions, wind-wave induced, and ocean surface current itself. The accurate wind-wave induced Doppler shift estimation is the major challenge for ocean surface current retrieval. All these issues need to be further investigated.
In the first half of this study, we use the most recent GDP (Global Drifter Program, GDP) drifter datasets from January 1st 2016 to December 31st 2020, which covers 6416 trajectories and more than 10 million observations, to investigate the contribution of geostrophic current to the total ocean surface current. The measured drifter velocity is the addition of the 15 meters depth large-scale current, the upper-ocean wind-driven current, the influence of tides and Stokes drifter and other forces on the drogue, and wind-induced slippage contribution. To account for the slippage contribution, a downwind velocity modeled as αW_s is subtracted from the drifter velocity, where W_s is 10-m wind speed, and α is the fraction of W_s converted to the slippage. After the slippage correction, we make a statistical comparison analysis between drifter velocity components, zonal (ut) and meridional (vt) velocity and the AVISO (Archiving Validation and Interpolation of Satellite Oceanographic Data, AVISO) geostrophic current, where the zonal (ugeo) and meridional (vgeo) velocities of geostrophic current are interpolated to the location of individual drifter observation by a three-dimensional time-space interpolation.
We carry out a direct comparison of the zonal velocity and meridional components between corrected drifter data and geostrophic current at a global scale, and calculate the probability in the cases of ugeo/ut < 0, vgeo/vt< 0. If ugeo/ut < 0 (vgeo/vt < 0), that indicates the ugeo (vgeo) is not the dominant contribution in the ut (vt), and the other ocean current processes dominate. Our preliminary results show that the portion of ugeo/ut < 0 reaches a value as large as 27.63% with a total number of effective observations is around 9.6 million, while the portion of vgeo/vt < 0 can be 31.59%. Additionally, while binning the corrected drifter data and the geostrophic current by a 0.5° × 0.5° bin, the portion of ugeo/ut < 0 and vgeo/vt < 0 could be 19.87% and 31.30%, respectively.
In the second half of this study, we investigate the effects of different wave directional spectra on the estimates of the wind-wave induced Doppler shift via a numerical Doppler model. By comparisons with some existing empirical or semi-empirical Doppler model, like CDOP model, KaDOP model, and KaDS model, our results show that the accurate estimation of wind-wave induced Doppler shift is highly dependent on the selection of wave spectra parameterization and directional spreading function. To accurately retrieve ocean current, a clear solution is to measure simultaneous and necessary ocean wave properties and wind vector to estimate the contribution that wind-wave induced.

The Copernicus Climate Change Service (C3S) is one of the six thematic services of the EU-funded Earth Observation programme Copernicus, managed by the European Commission (EU). C3S is implemented by the European Centre for Medium-Range Weather Forecasts (ECMWF), and it has as primary objective providing access to authoritative climate information in support of climate adaptation and mitigation policies of the EU.
C3S supports the implementation needs of the Global Climate Observing System (GCOS) and in turn, the objectives of the United Nations Framework Convention on Climate Change (UNFCCC), by assuring timely access to a large number of quality assured Climate Data Records (CDRs) of Essential Climate Variables (ECVs) derived from space-based Earth observations. In total, GCOS specifies a total of 54 ECVs which are critical to characterize the climate system (relevant), measured globally with existing technologies (feasible) and at an affordable level of investment (cost-effective). C3S has already implemented services for 22 of these land, ocean and atmosphere ECVs. Target requirements for most of the CDRs of ECV products, in terms of uncertainty, stability, temporal and spatial resolution, are based on the framework defined in the GCOS 2016 Implementation Plan (GCOS-IP 2016). However, the ECV services implemented in C3S have proven that target requirements are not always attainable with the current technology, what in turn also provides a valuable feedback for the upcoming update of the 2022 GCOS-IP.
C3S has recently finalized the successful implementation of the first phase of ECV services with access to data and associated information products through the C3S Climate Data Store (CDS). The CDS offers open and free access to an evolving catalogue of climate data products about data of the past, present and future, as well as tools to enable their use. The CDS counts already more than 100,000 registered users. It currently offers access to ECV products associated with 22 ECVs. Each data product provides state-of-the-art reliable access to quality-assured and regularly updated CDRs and Interim CDRs with global or near-global coverage, as well as comprehensive supportive documentation. In addition, an independent evaluation of the data products and associated services assures their quality and roadmap towards target requirements. Also, the implementation of a series of online applications or data viewers provides simple examples of the use of the data accessible through the CDS for climate purposes.
In this presentation we will show the status of the C3S ECV services in the first year of Copernicus 2 (COP 2), we will provide an overview of their main individual components and the plans of these services for the whole COP 2 period, between 2021-2027.

Contribution to the operational monitoring of the climate and the detection of global climatic change is a major objective of EUMETSAT. The growing relevance of this objective can be sensed by the fact that our societies are now actively managing weather and climate risks as a core task. The first three risks identified in the 2020 World Economic Forum Risk Report are weather and climate-related. Extreme weather becoming more likely in a changing climate has been continuously identified as the biggest global risk in terms of likelihood and is in the top five in terms of impact. Failure to implement climate change mitigation actions and natural disasters are second and third-ranked risks respectively, both in terms of impact and likelihood.

Adaptation and mitigation strategies require a full combination of operational weather and climate information services built on solid scientific foundations, including observations, seasonal and decadal predictions, climate projections, climate analyses, and assessments of impacts. Mitigating climate change and adapting to the impacts of weather extremes and changing environmental conditions require detailed information on all relevant scales. Earth system observations from space are a key component of such strategies and provide valuable information for the benefit of society. EUMETSAT engages in providing satellite data, information, and services dedicated to a better understanding of climate variability and change including the characterisation of extreme events as well as impacts of these changes on society.

EUMETSAT and its network of Satellite Application Facilities have released close to 100 climate data records since starting sustained climate monitoring activities in 2008. Most Climate Data Records were generated by reprocessing of the original sensor data, covering long time-series (~40 years) and several generations of instruments followed by the creation of GCOS Essential Climate Variables from the sensor data. According to the ECV Inventory of the joint CEOS-CGMS Working Group on Climate, the data records produced by EUMETSAT providing inputs to 19 out of 37 ECVs rated observable from space. Data from EUMETSAT are used in many national climate services and the European Copernicus Climate Change Service that EUMETSAT specifically supports with inputs for global and regional reanalysis and the ECV data sets generated by its Satellite Application Facility Network. The usage of the data for climate change analysis is documented in the recent IPCC report on the physical basis as well.

This paper presents an overview of the EUMETSAT’s strategy for the future development of its support to climate services addressing the usage of the evolving space segment, the development of new products from rescued past data, and research to operations and operations to research exchanges involving new cloud infrastructures.

The major challenge for climate data is the integration of benefits of legacy and new generation systems. This involves EUMETSAT systems and third party system of the past, e.g., NOAA instruments on Metop satellites. With the launch of MTG and EPS-SG more data records at instrument level will be generated and maintained. For instruments with a heritage it means addition to the past and improvement of the past data. In addition, it means integration with similar instruments in different orbits, e.g., EPS-SG MetImage with NOAA VIRS and past AVHRR. For new sensors, such as MTG IRS it means new data records. The new instruments hold a strong capability for new products, e.g., MWI and ICI hold potential for precipitation climatology but need to be used together with NASA GPM, geostationary data and the heritage of microwave radiometers to build a climatology. This is a huge task but Europe still lacks a global precipitation climatology using this data. In addition, information on aerosols and clouds will be improved using the 3MI instrument on EPS-SG, but how to utilise this for climate monitoring needs to be studied.

Sentinel satellites operated by EUMETSAT on behalf of the EU have very high relevance for climate change monitoring as well. In particular, Sentinel-6 Michael Freilich altimeter sea-level estimates and the planned CO2M mission with CO2 and CH4 estimates are highly relevant and are included in EUMETSAT’s framework in support of climate services. In addition, the development of an understanding on how new capabilities such as the planned microwave sounding constellation and Doppler Wind LIDAR can be utilised for climate monitoring. Similarly, it is important to continue to take into account specific climate requirements in the planning and definition of future EUMETSAT mandatory programmes such as M4G and EPS-TG.

Coordination and cooperation with other space agencies is a must, e.g., the utilisation of past, current, and future geostationary observations for climate monitoring is a challenge as up to 50 geostationary satellite missions operated by many agencies are part of the record with a variety of instrumentation since the late 1970s. Data from the geostationary ‘ring’ are essential for the provision of many global GCOS ECV products that are used in climate service applications and climate change research. EUMETSAT is engaged in data rescue, uncertainty characterisation, recalibration, harmonisation and reprocessing of such data and aims at the provision of the geostationary ‘ring’ data to users in the near future on its joint EUMETSAT-ECMWF cloud infrastructure the so called European Weather Cloud and the EUMETSAT Data Store.

Further improvements of the EUMETSAT climate data portfolio aim at the inclusion of new products developed by European research. EUMETSAT will help bridge the gap between research and operations by importing scientific methods and exporting the infrastructure for processing, operational data management and the distribution mechanism using the European Weather Cloud. The cloud infrastructure also offers new opportunities to tailor data products to uses to deliver records of information complementing the data records. In addition, investments will be made to improve the EUMETSAT Data Store and to strengthen the user engagement by stimulation measures, e.g., via the EUMETSAT user training programme.

Vector-borne diseases (VBD) account for about 17% of all lost life, illness, and disability globally. These diseases are gradually expanding their spatial range emerging in regions where they were typically absent, and re-emerging in areas where they had subsided for decades. They apply pressures that stretch scarce resources to breaking point. Climate change is expected to make these diseases more widespread, frequent and unpredictable. Dengue is the fastest spreading mosquito-borne viral disease in the world. About half of the global population lives at risk of the disease. Dengue is disproportionately linked to poverty and inequality in many low and middle-income countries.A fundamental gap is that current dengue control strategies are essentially reactive, meaning that they take place after cases occurred. These reactive responses reduce the probability to control and mitigate outbreaks. Disease models driven by Earth observations and seasonal climate forecasts could provide useful information of future disease risk to allow early action. We developed a superensemble of probabilistic models to predict dengue incidence across Vietnam up to 6 months ahead at a sub-national scale. The modelling framework was codesigned with stakeholders from the World Health Organization, the United Nations Development Programme, the Vietnamese Ministry of Health, the Pasteur Institute Ho Chi Minh City, the Pasteur Institute Nha Trang, the Institute of Hygiene and Epidemiology Tay Nguyen, and the National Institute of Hygiene and Epidemiology. The model superensemble generated accurate predictions evaluated using proper scores. The skill of the model varied with geographic location, forecast horizon, and time of the year, performing at its best during the peak season and in areas with high levels of transmission. Evaluated using a theoretical cost-loss analysis, the superensemble had considerable value in most provinces relative to not using the forecasting system. The system is being prospectively evaluated and the framework is being extended to other areas of the world.

In situ observations of 2m air temperature (T2m) are widely used in climate science and services. Land surface temperature (LST) is strongly related to T2m, and can be observed using InfraRed (IR) or MicroWave (MW) satellite observations, with several satellite LST datasets now freely available to users. Long-term satellite LST data are desirable for climate science and services because they can provide temporal and spatial information on surface temperatures that cannot be achieved using in situ T2m. In particular, these data can offer global coverage, enabling better understanding of regional climate change and impacts in areas with sparse in-situ data. Additionally, satellite LST data provides an independent measure of surface temperature change and for some applications, additional information not available through T2m observations. For example, over dense vegetation, LST represents the canopy temperature and could therefore be used more directly than T2m to indicate vegetative water status. Over vegetation-free regimes, LST represents the temperature of the land surface, and could be used to monitor road surface temperatures more accurately than T2m. However, studies of temperature change require the satellite LST data to be free from non-climatic discontinuities, which is not always the case, with previous studies showing variable agreement between LST and T2m time series. The European Space Agency’s Climate Change Initiative for LST (LST_cci) aims to deliver a significant improvement to the capability of current satellite LST data records to meet the Global Climate Observing System requirements for climate applications and realise the full potential of long-term LST data for climate science and services. The objective of this study is to assess the stability and trends in six new LST_cci datasets and demonstrate the value in using these LST data to augment T2m observations.
LST anomalies are compared with homogenized station T2m mean, minimum and maximum anomalies over Europe, which verifies that the LSTs in all six of these datasets are well coupled with T2m (LST vs T2m anomaly correlations and slopes range between ~0.6 and ~0.9). Having confirmed the relationship between the LST and T2m anomalies, the homogenized T2m data are used to assess the temporal stability of the LST_cci data through a comparison of the LST and T2m anomaly time series. Only the LST_cci datasets for the MODerate resolution Imaging Spectroradiometer (MODIS) on-board Aqua and the Advanced Along-Track Scanning Radiometer (AATSR) appear stable; the LST_cci MODIS/Terra, ATSR-2, and multisensor InfraRed and MicroWave datasets show non-climatic discontinuities associated with changes in sensor and/or drift over time. For MODIS/Aqua (2002-2018), statistically significant trends in LST of 0.64-0.66 K/decade are obtained, which compare well with the equivalent T2m trends of 0.52-0.59 K/decade; there is no statistically significant difference between these trends in LST and T2m. The LST and T2m trends for AATSR (2002-2012) are found to be statistically insignificant, likely due to the comparatively short analysis period and specific years available for analysis.
In addition, this study demonstrates that the way in which LST and T2m data are compared can affect results. The IR satellite LST data are only available for cloud-free conditions and when cloudy T2m anomalies are excluded from the comparison, a ‘clear-sky bias’ is introduced. This results in an annual cycle and non-zero mean difference in the monthly mean LST-minus-T2m anomaly time series, which are both removed when the monthly T2m anomaly time series is regenerated from ‘all-sky’ T2m observations. Nevertheless, when the trends in T2m are recalculated from the ‘all-sky’ T2m data, they are still statistically significant and almost identical (0.01-0.06 K/decade difference) for the MODIS/Aqua time period. Therefore, it is concluded that the results presented in this study do not offer any evidence that a clear-sky bias affects trends calculated using cloud-free IR observations. The importance of having long and homogeneous time series for climate applications, including LST, is also highlighted in this study through a comparison of trends for T2m for the MODIS/Aqua and MODIS/Terra time periods. Here, the addition of two additional years of data at the beginning of the MODIS/Terra period (2000-2001) is found to reduce the trends in T2m by ~0.2 K/decade.
This study suggests that satellite LST data have great potential to be used in climate science and services. In particular, the analysis presented here demonstrates that satellite LSTs can be used to augment land T2m data where time series of surface temperatures are required, provided the required homogeneity is assured.

The Ocean Colour Climate Change Initiative (OC-CCI) project has spent a decade working to create an ever-improving climate data record (CDR) of the ocean colour Essential Climate Variable (ECV). The cyclical process of improvement, release and feedback has led to 6 versions of the dataset being created to-date with incorporation of data from additional sensors, incremental improvements at all stages of the processing chain, from top-of-atmosphere derived pixel flagging to inter-sensor bias correction and blended in-water algorithms. The OC-CCI dataset is the highest volume dataset of the ESA ECV catalogue, containing over 90 variables, spanning 23 years and now being provided for testing purposes at 1km resolution. The OC-CCI dataset has been used in over 180 publications including a number of prominent climate related studies such as the upcoming IPCC working group 2 report and publications in eminent journals such as Nature (Tang et al. 2021, Dutkiewicz et al. 2019). The data has been used to study the impacts of climate from the local scale, such as climate impacts on Sea Urchin Settlement (Okamoto et al. 2019), to the global scale, such as computation of oceanic primary production (Kulk et al. 2020). The data have also been used to investigate a variety of cross-disciplinary research areas such as the connections between ocean physics and biology (Balaguru et al. 2018), the synergy between ocean models and remote-sensing observations (Baird et al. 2020), and the links between ocean and human health (Campbell et al. 2020). This uptake of the data by the scientific community has been aided by the standardised formatting and multiple channels for access to the data (from bulk downloads to web-based interactive browsers). The processing chains that were developed in a research phase under OC-CCI are now used for operational data processing for services such as CMEMS and C3S. Here, we reflect on the lessons learned over the last decade of ECV development and consider the current maturity and future of the OC-CCI dataset in terms of 1) transforming research mode data into operational products, 2) data accessibility and interface with toolboxes, 3) the percolation of the data into decision-making spheres, and 4) the use of the data in cross-disciplinary science.


References:
Baird, M; Chai, F; Ciavatta, S; Dutkiewicz, S; Edwards, C; Evers-King, H; Friedrichs, M; Frolov, S; Gehlen, Ma; Henson, S; Hickman, A; Jahn, O; Jones, E; Kaufman, D; Mélin, F; Mouw, C; Muhling, B; Rousseaux, C; Shulman, I; Wiggert, J. Synergy between Ocean Colour and Biogeochemical/Ecosystem Models. IOCCG Report Number 19 (2020).

Balaguru K, Doney SC, Bianucci L, Rasch PJ, Leung LR, Yoon J-H, et al. (2018) Linking deep convection and phytoplankton blooms in the northern Labrador Sea in a changing climate. PLoS ONE 13(1): e0191509. https://doi.org/10.1371/journal.pone.0191509

Campbell AM, Racault MF, Goult S, Laurenson A. Cholera Risk: A Machine Learning Approach Applied to Essential Climate Variables. Int J Environ Res Public Health. 2020;17(24):9378. Published 2020 Dec 15. doi:10.3390/ijerph17249378

Dutkiewicz, S., Hickman, A.E., Jahn, O., Henson, S., Beaulieu, C. and Monier, E., 2019. Ocean colour signature of climate change. Nature communications, 10(1), p.578.

Tang, W., Llort, J., Weis, J. et al. Widespread phytoplankton blooms triggered by 2019–2020 Australian wildfires. Nature 597, 370–375 (2021). https://doi.org/10.1038/s41586-
021-03805-8

Kulk, Gemma & Platt, Trevor & Dingle, James & Jackson, Thomas & Jönsson, Bror & Bouman, Heather & Babin, Marcel & Brewin, Bob & Doblin, Martina & Estrada, Marta & Figueiras, F.G. & Furuya, Ken & González-Benítez, Natalia & Gudfinnsson, Hafsteinn & Gudmundsson, Kristinn & Huang, Bangqin & Isada, Tomonori & Kovač, Žarko & Lutz, Vivian & Sathyendranath, Shubha. (2020). Primary Production, an Index of Climate Change in the Ocean: Satellite-Based Estimates over Two Decades. Remote Sensing. 12. 826. https://doi.org/10.3390/rs12050826

Daniel K. Okamoto, Stephen Schroeter, Daniel C. Reed (2019) Effects of Ocean Climate on Spatiotemporal Variation in Sea Urchin Settlement and Recruitment, bioRxiv 387282; doi: https://doi.org/10.1101/387282

The seriousness of the climate emergency and the need for substantial and rapid action is increasingly widely recognised. However, many key actors are stuck, not knowing how to respond. The UCL Climate Action Unit works with communities of practice, communities of and place, institutions, policy makers, and the general public to assist them in identifying their agency to act. Space EO data can provide key information to guide such actions. Our theory of change addresses the behavioural, social, and political obstacles which get in the way. It integrates scientific insights from neuroscience, psychology and the social sciences into the delivery of real-world interventions. We engage with professionals and experts in public, private and civil society organisations to help them develop more effective responses to climate change. We have a growing portfolio of examples of improving the way climate risk information drives decision making in policy and business, equipping experts and communicators to tell better and more varied stories to better motivate action, and helping key communities and organisations to deliver climate-positive action. The key is that awareness and information alone do drive action ‘Actions Drive Beliefs”. The presentation will summarise the underlying principles, provide examples of recent successes and show the relevance to the exploitation of space EO data and services.

REACT is Skytek's main commercial product, a SaaS platform that addresses the accumulation challenges of underwriting insurance risks.
REACT is a web-based system that uses real-time information to create accurate insights and actionable decision-ready data to support insurance and brokers in making smarter, faster, and better decisions. REACT provides the ability to integrate and analyse a variety of datasets, combining geolocation datasets (e.g. AIS), satellite imagery (optical and radar), big data, and machine learning, enabling monitoring of fixed and moving assets, including property, vessels and energy assets in real time, clearly identifying clusters, risks, and aggregated exposure.
REACT’s main features provide capabilities for:
- Global Risk Aggregation estimation, in a dynamic way and in real time, for any asset and in any location, with automated calculation of global exposure and ranking such as top 5 risk aggregation locations
- Aggregation can be analysed at portfolio or region level, with fully configurable threshold levels to suit the required outcome
- Severe Weather tool, allowing linkage with real time weather reports to monitor pre-event exposure aggregation and therefore support additional reinsurance purchase. Within this tool it is possible to analyse historical patterns of windstorm behaviours and locations affected
- Port Cargo accumulation information, allowing near real time cargo analysis, using Earth Observation imagery. We apply machine learning proprietary algorithms to count containers, refrigerated containers and cars at a port
- Automated and fully configurable alerts for monitored regions and assets
REACT is a powerful system that can be used to support:
- Modelling systems, as effectively complimentary tool
- Underwriting processes, for easy pre-screening of new clients, claims historical analysis, risk scoring and comparing
- Sanction compliance analysis and reporting, including War risk breach assessment
- ESG compliance and reporting, specifically in terms of environmental scores, sustainability and decent employment compliance
We plan to present how REACT has supported situations such as the
- Ever Given Grounded-Suez Canal, providing in real time the full list of assets in affected area, preventing loss creep and providing insurance with the ability to quickly report on potential exposure
- Several hurricane pre/post analysis examples, using Earth Observations data for damage assessment and providing insurances with an accurate and fast was to detect and view damage to individual properties within a portfolio, highlighting the storm path and providing supporting information on infrastructure damage.

Disaster Risk Financing (DRF) represents an effective risk management option, able to complement more traditional Disaster Risk Reduction measures by offering a lean way to unlock resources for a fast and prompt reaction when a disaster occurs. This is particularly important when sovereign risk is considered in countries where the magnitude of flood events often overcomes the response capacity of the institutions. A typical example is South East Asia where extensive monsoon flooding often endures for months and hits large geographical areas.
In this context, recent improvements in EO Products allowed the development of enhanced detection capacity in support of such applications. One example are the services developed by the project e-Drift (Disaster RIsk Financing and Transfer), financed by ESA and led by CIMA Research Foundation (IT).
The project is framed in the cooperation between ESA and the World Bank/DRF Initiative that provided not only guidance for the service development, but also created the preconditions for its operational use in the context of SEADRIF : a regional platform that provides participating nations in Southeast Asia with advisory and financial services to increase preparedness, resilience and cooperation in response to climate and disaster risks.
eDrift created a processing environment that enables access to seamless EO added value services in support of DRF applications.
In particular, the platform allows the continuous monitoring of an extended Region of Interest (e.g. a full country) taking advantage of the Sentinel-1 constellation global acquisition capability and of the free and long-term access to data guaranteed by the EU Copernicus services. A flood extent map can be updated every day by making use of any Sentinel-1 acquisition that intersects the region of interest, guaranteeing the best possible coverage. The process is fully automated and guarantees a high rate of data fetching.
Thanks to its reliability and high degree of automation the service does not require supervised intervention. Therefore, the same service can also be used to create an exhaustive set of past flood events on the analyzed area by processing the full archive of acquired satellite images. Such a product can be translated into historical flood frequency maps that are used to calibrate catastrophe models in support of insurance applications.
A first application that has been tested is parametric insurance to unlock resources for immediate and effective response in the aftermath of a flood event.
The eDrift services are fed into a platform supporting SEADRIF in the delivery of a parametric insurance product.
In this platform the EO services provided by eDrift are combined with flood models to determine the number of affected people at country level. This output can be used to activate a parametric insurance in support of the countries for response operations.
This service has been operational since February 2021. The EO component of the service is provided by the eDrift and the EO flood mapping is the first operational service on the market from the eDrift service portfolio.
Within the same period the eDrift project has been working to address challenges evidenced by the users especially in urban areas where the flood detection capabilities of synthetic aperture radar (SAR) sensors remains limited. Cutting edge algorithms for urban flood detection have been tested in several case studies and their operational use will soon be tested to enhance these types of services in support of DRF.

Soil Moisture Index Insurance Mozambique - protecting smallholder farmers against drought.

The right soil moisture conditions are essential to obtain good yield for farmers around the world, especially to those who farm in rainfed conditions. In Mozambique, agriculture provides livelihoods to almost 81 percent of the population (World Bank, 2015). The majority of crop production is undertaken by smallholder families who grow staple crops for consumption. These farming families face an increased likelihood of extreme weather events such as flood, drought, and cyclones due to climate change. Currently, only 34,378 of smallholder farmers in Mozambique have at least insurance against extreme climatic risks (MADER, 2021). This represents less than 1% of the total population of more than 4 million smallholder farmers in the country, as per the estimations contained in the most recent survey of the Ministry of Agriculture (FSDMoç, 2021).

Hollard Insurance is a pioneering insurance company attempting to de-risk Mozambican smallholder farmers by developing and delivering agricultural insurance products. In collaboration with Phoenix Seeds and a number of development agencies and national agricultural development projects, Hollard insurance provides farmers with improved crop seed varieties, bundled with satellite-based index insurance. Mozambican farmers who secure their inputs from the covered entities also receive reassurance that should a drought event occur, current and future growing seasons will not be a complete loss. Developing trust in the product among agrarian communities with low levels of insurance awareness is crucial for Hollard Insurance. They need to rely on a parametric insurance product which covers the actual crop damage on the ground, and is affordable and easy to explain to farmers.

The insurance product which covers farmers in the current growing season (nov ‘21 - april ‘22) is developed by reinsurance firm SwissRe using satellite-derived soil moisture data provided by VanderSat. The soil moisture measurements are based on passive microwave observations, which due to their unique sensor characteristics allow for all-weather retrievals and therefore do not have cloud cover issues as optical based EO data. VanderSat developed a unique patented technology (EU Patent # 17 728 899.0) that allows downscaling brightness temperatures to field level scale (100m), which are then transformed to soil moisture using the Land Parameter Retrieval Model (Owe et al., 2008; Van der Schalie et al., 2017). This is a large step forward from other state-of-the-art soil moisture data sets (e.g. Dorigo et al., 2017; Chan et al., 2018), which have a resolution of 9km at best. Soil moisture indices reflect directly the water content in the soil which is available for the plant to grow, and therefore correlate highly with obtained crop yields in dominantly rainfed farming conditions compared to vegetation and weather-derived indices (Mladenova et al. , 2017; Vergopolan et al., 2021). Soil Moisture data at field level with a long historical record (20 years, by using AMSR-E, AMSR2 and SMAP) enables the development of scalable index-based insurance products.

The drought insurance product covers the seeding, vegetative and harvesting phases of plant growth and is tuned to the drought sensitivities of the insured crop. For each growing phase, different drought triggers are developed using parametric rating (SwissRe, 2021). The satellite tracks developing drought conditions during the season, and a payout is triggered when a drought becomes severe. The soil moisture values and payouts can be accessed through an online dashboard which improves transparency in the insurance value chain. Drought is the most important peril that inland farmers are facing in Mozambique. Farmers in the coastal regions are frequently plagued with tropical cyclones that wipe out harvests. Ongoing research and development activities are performed to best capture the impact of cyclones on coastal agricultural land using parametric index insurance.

References

World Bank (2015), Mozambique: Agricultural Sector Risk Assessment. Risk Prioritization, accessible via: https://openknowledge.worldbank.org/handle/10986/22748

Ministério da Agricultura e Desenvolvimento Rural (MADER) (2021), Inquérito Agrário Integrado 2020, Maputo – MADER, accessible via: https://www.agricultura.gov.mz/wp-content/uploads/2021/06/MADER_Inquerito_Agrario_2020.pdf

Financial Sector Deepening Mozambique (FSDMoç) (2021), Mozambique Inclusive Insurance Landscape Report 2021, accessible via:
http://fsdmoc.com/news/launching-report-landscape-inclusive-insurance-mozambique/

Owe, M., de Jeu, R., & Holmes, T. (2008). Multisensor historical climatology of satellite‐derived global land surface moisture. Journal of Geophysical Research: Earth Surface, 113(F1).

Van der Schalie, R., de Jeu, R. A., Kerr, Y. H., Wigneron, J. P., Rodríguez-Fernández, N. J., Al-Yaari, A., ... & Drusch, M. (2017). The merging of radiative transfer based surface soil moisture data from SMOS and AMSR-E. Remote Sensing of Environment, 189, 180-193.

Dorigo, W., Wagner, W., Albergel, C., Albrecht, F., Balsamo, G., Brocca, L., ... & Lecomte, P. (2017). ESA CCI Soil Moisture for improved Earth system understanding: State-of-the art and future directions. Remote Sensing of Environment, 203, 185-215.

Chan, S. K., Bindlish, R., O'Neill, P., Jackson, T., Njoku, E., Dunbar, S., ... & Kerr, Y. (2018). Development and assessment of the SMAP enhanced passive soil moisture product. Remote sensing of environment, 204, 931-941.

Mladenova, I. E., Bolten, J. D., Crow, W. T., Anderson, M. C., Hain, C. R., Johnson, D. M., & Mueller, R. (2017). Intercomparison of soil moisture, evaporative stress, and vegetation indices for estimating corn and soybean yields over the US. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 10(4), 1328-1343.

Vergopolan, N., Xiong, S., Estes, L., Wanders, N., Chaney, N. W., Wood, E. F., ... & Sheffield, J. (2021). Field-scale soil moisture bridges the spatial-scale gap between drought monitoring and agricultural yields. Hydrology and Earth System Sciences, 25(4), 1827-1847.

SwissRe (2021), Drought is insurable, accessible via: https://www.swissre.com/risk-knowledge/mitigating-climate-risk/natcat-2019/drought-is-insurable.html

Many parametric or index-based drought risk financing instruments are based on satellite-derived rainfall, temperature and/or vegetation health data. However, an underlying issue is that indices often do not perfectly correlate to the actual losses experienced by the policy holders. The resulting increased basis risk can diminish demand for parametric drought risk insurance. Remotely sensed soil moisture (SM) can help decrease basis risk in parametric drought insurance through 1) complementary and/or improved parameters and variables used in existing models such as the Water Requirement Satisfaction Index (WRSI), 2) shadow-models to cross-check, test or validate payouts triggered through other indicators and models or 3) potentially through development of a stand-alone product. Here, we will demonstrate the use of a combined Sentinel-1 and Metop ASCAT high resolution soil moisture dataset to predict yield and develop an early-warning yield deficiency indicator for Senegal.

Soil moisture is retrieved from the Advanced SCATterometers (ASCAT) on-board the Metop satellite series, which has an original spatial sampling of 12.5 km. Sentinel-1 backscatter data at 500 m spatial sampling is used to downscale the Metop ASCAT surface soil moisture data to 500 m. The underlying concept is the temporal stability of surface soil moisture: In the temporal domain surface soil moisture measured at specific locations is correlated to the surface soil moisture content of neighbouring areas, where neighbours with similar physical properties (like soil texture, land cover and terrain) show a higher coherence to the local surface soil moisture than others. In addition to soil moisture, freely available rainfall from Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) and Copernicus Global Land Service NDVI were used. All datasets were spatially resampled to a 500 m grid, temporally aggregated to monthly anomalies and finally detrended and standardized. Data on yields was obtained from the Food and Agriculture Organization of the United Nations (FAO). Data on crop growth areas is based on FAO Global Agro-Ecological Zones (GAEZ) information and Livelihood zones (2015).

First, regression analysis with yearly yield data was performed per EO dataset for single months. The EO datasets were aggregated over areas where the specific crop was grown. Secondly, based on these results multiple linear regression was performed using the months and variables with the highest explanatory power. The multiple linear regression was used to provide spatially varying yield predictions by trading time for space. The spatial predictions were validated using sub-national yield data from Senegal and reports from the African Risk Capacity (ARC).

The analysis demonstrates the added-value of satellite soil moisture for early yield prediction. Soil moisture showed a high predictive skill early in the growing season: negative early season soil moisture anomalies often lead to lower yields. NDVI showed more predictive power later in the growing season. Combining anomalies of the optimal months based on the different variables in multiple linear regression improved yield prediction. Especially at the start of the season soil moisture improves predictions, with the ability to explain 60% (groundnut), 63% (millet), 76% (sorghum) and 67% (maize) of yield variability. These findings are particularly relevant for parametric drought insurance, because an earlier detection of drought conditions enables earlier payouts, which then help to mitigate the development of shocks into serious crises with often long-lasting socioeconomic effects.

Based on the analysis a yield deficiency indicator can be developed, which can provide spatial information on yield deficiencies. Yield deficiencies were compared to sub-national yield information and WRSI information as reported by the African Risk Capacity end of season reports. Strong spatial correspondence was found between the yield deficiency indicator and WRSI. For example, for millet in Senegal for the drought 2019 strong yield deficiencies in the provinces of Ziguinchor, Fattick, Kaolack and Kaffrine and moderate deficiencies in Thies, Louga and Tambacounda were found. This corresponded to low WRSI as reported by the African Risk Capacity in its end of season report of 2019. The analysis shows very clearly that soil moisture can be a valuable tool for anticipatory drought risk financing and early warning systems.

This analysis was performed in collaboration with the World Bank Disaster Risk and Financing Program and Global Risk Financing Facility.

The economic impacts of floods push people into poverty and cause setbacks to development as government budgets are stretched and people without financial protection are forced to sell assets. Investments in flood mitigation and adaptation, such as expanding insurance coverage through index based insurance, i.e. direct payouts based on predetermined indexes of e.g. flooded area, could reduce anticipated losses from floods and increase resilience. Insurance penetration remains low ( < 1%) for climate-vulnerable populations in countries like Bangladesh, which urgently need financial protection from extreme floods to protect development. Expanding insurance coverage requires the ability to quantify flood risk and monitor it in near real-time in remote locations, which is challenging due to limited data in most areas. Satellite observations have the potential to fill this data gap and expand insurance coverage by providing globally available observations available at regular intervals.

Algorithms to map flood extent are improving by leveraging machine learning and multiple sensors. One way to estimate the frequency of extreme floods is to measure the spatial extent of inundation directly from space. Given the growing length of the satellite record, time series of inundated area could be used for exceedance probability estimation to develop insurance products, but relies on the availability of the data over extended periods of time ( > 30 years).

High spatial resolution satellites (such as Sentinel-1 and 2, 10m) have not been around long enough for reliable exceedance probability estimates to design index insurance triggers.Using deep learning, we fuse the daily MODIS time series with Sentinel-1 and Harmonised Landsat Sentinel-2 (HLS) data to create 20 year historical inundated area estimates over Bangladesh.
We benchmark consistency of the time series against in situ records of 42 water level stations for validation. Satellite fusion could generate longer time series of inundation, to eventually generate return period estimates at watershed or country scale beyond Bangladesh.

Agricultural production inherits comparably high production risks and weather-related crop yield losses are even expected to increase due to climate change. The government of India is supporting risk mitigation of Indian farmers for crop loss caused by natural disasters through the Pradhan Mantri Fasal Bima Yojana (PMFBY) launched in 2016. PMFBY aims to support sustainable production in the agricultural sector. PMFBY supersedes the earlier insurance schemes National Agriculture Insurance Scheme (NAIS), Weather-based Crop Insurance scheme and Modified National Agricultural Insurance Scheme (MNAIS). The crop insurance coverage under PMFBY increased substantially since its inception, reaching up to 50% (in 2018-19), with a governmental contribution in the fiscal year 2021-22 of approx. 1.8 billion EUR. PMFBY requires the enhanced usage of technology for the reporting to the National Crop Insurance Program (NCIP) Portal and encourages the move towards increasing use of remote sensing and Artificial Intelligence/ Machine Learning methods.
The overall aim of the India Crop Monitoring Project, a cooperation between Munich Re, Potsdam Institute for Climate Impact Research (PIK) and GAF AG, is to support direct agro-insurers in India with information about historical and current state of agricultural land through a combination of Earth Observation and weather information products. This information is retrievable via the web application AgroView®.
AgroView® is a single-page web GIS application, using a frontend based on HTML5, JavaScript and CSS technologies. The backend contains a spring framework with REST services, Java and Tomcat; PostgreSQL/ PostGIS are employed as a database. OGC-compliant geodata services, such as WMS, are provided via GeoServer. The whole application and its components are implemented in a Kubernetes Cluster in the Open Telekom Cloud (OTC) environment, offering a scalable, high-performance processing environment, with direct access to various Earth Observation Data Archives.
AgroView® integrates, manages and displays geographical raster, vector and alphanumerical information from various sources. Besides integration of data from external sources (e.g. administrative boundaries, weather data, historic events), historic and current satellite data is integrated and analysed (e.g. Sentinel-1 and Sentinel-2, CHIRPS, MODIS). The data backend of the applications gathers satellite images and weather data from the providers’ archives, processes them into relevant crop health and weather condition/ drought indicators in a fully automated processing chain, and provides the produced information products to the User in near-real time. The analysis functionality of the application allows the exploitation of the multitude of data at different scales from national to field level, and throughout time at full temporal resolution of the source data from historic to current.
With its wide range of functionality, AgroView® supports the User across the full Crop Insurance Cycle, and its key features all the User to reduce risks throughout the different phases of the growing season efficiently, and increase profitability. In the insurance tendering phase, the User can analyse historic weather patterns and agricultural variations for risk analysis and pricing. Real-time and historical crop field information helps to understand land use in support of decision of enrolment areas in support of insurance coverage and expansion. During the sowing period, the User can monitor rainfall and soil moisture to derive probability of crop failure, and help to invoke prevented sowing and the assessment of correspondent triggers.
Throughout the growing season, AgroView® provides crop health monitoring through a set of vegetation indices, and the User is alerted when crop condition indicates agricultural drought. The continuous monitoring of weather data allows the timely detection of mid-season adversities such as floods, prolonged dry spells or severe droughts.
In addition to this continuous monitoring of vegetation and rainfall conditions through the application, the User provided with fortnightly reports on his portfolio areas in PDF format through the fully automated reporting function of AgroView®. This automated reporting allows the User to make informed decisions in a timely and efficient manner.
During harvest period, the application supports post-harvest loss assessments by combining the monitoring of harvesting patterns through vegetation indices in combination with real time rainfall data tracking and identification of peak rainfall periods. With its integrated Yield Estimation System (YES), the User benefits from predictive crop yield modelling through the DSSAT model. Finally, the User can plan his crop cutting experiments for yield assessment based on the predicted yields for major crops and NDVI data, for reporting into the PMFBY system.
In summary, the stakeholders benefit from full adaptability to their specific business requirements in the insurance sector, and fast integration of dedicated geo-technological data analytics fully scalable from national to field level. The application provides the agro-insurance User a wide range of built-in tools and functionalities, adjusted to his needs in the PMFBY system, in a User-friendly package through easy browser access. It therefore constitutes a fully-fledged, tailored digital farming solution for insurance in India.

Bandung is a capital city of West Java Province, one of the largest city by population in Indonesia. The city experience rapid development of industrialization since 1970s. The growth of population increase afterwards, and create demand of resources to support the city, including fresh water. Ground water has become inevitable source to fill the high demand while on the other hand, government cannot meet the demand of fresh water. This supply gap causes exploitation of ground water, and triggered environment quality degradation, which can be indicated by high land subsidence rate. We processed ALOS-PALSAR 1 data in the period of 2006-2011, then Sentinel-1 for period of 2014-2020. Small Baseline Subset Analysis (SBAS) is used for time-series analysis. GPS annual campaign were also done for period of 2005-2018, and we also manage to put several low-cost GPS stations in some prominent points to calibrate InSAR data processing. Our results shows that between 2006-2011, the area of land subsidence mostly located on the industrial area, with the highest rate up to 13 cm/year in Leuwi Gajah Industrial Complex. Some other areas which suffer land subsidence during the same period, are also located in the industrial complex such as Gede Bage, Ranca Ekek, and Majalaya. However, the spatial distribution of land subsidence is shifting during the period of 2011-2015. The demand of ground water from household grows rapidly within this period, indicated by a much higher rate of land subsidence occue in housing area. The area and rate of land subsidence broaden significantly in some housing areas such as Kopo and Margahayu with subsidence rate of 12 cm/year and 8 cm/year respectively. The opposite trend happened in the previously stated area, slowing down of land subsidence rate are observed in several locations with 5-8 cm/year slower compare to 2006-2011 period. The impact of land subsidence in housing area has caused the widen of flooding area and degradation of ground water quality, especially in Kopo District which suffers highest subsidence rate in housing area.

Many surface and deep aquifers in Central Mexico are overexploited to address water needs for public/private and industrial use, with boosting demand in expanding cities and metropolises. As a consequence, aquifers deplete and urban centers are widely affected by land subsidence and its derived risk for housing, transport and other infrastructure, with associated economic loss, thus representing a key topic of concern for inhabitants, authorities and stakeholders. This work analyses groundwater resource availability and aquifer storage change in Central Mexico based on groundwater management reports, issued by the National Water Commission and published yearly in Mexico's Official Federal Gazette. Evidence from piezometric measurements and aquifer modeling are discussed in relation with satellite InSAR surveys based on geospatial analysis of Sentinel-1 IW SAR big data processed within ESA’s Geohazards Exploitation Platform (GEP), using the on-demand Parallel Small BAseline Subset (P-SBAS) service. Case studies include a number of aquifers where significant groundwater deficits and aquifer storage changes were estimated over the last years, including those in the Metropolitan Area of Mexico City [1], one of the fastest sinking cities globally (up to 40 cm/year subsidence rates); the state of Aguascalientes [2], where a structurally-controlled fast subsidence process (over 10 cm/year rates) affects the namesake valley and capital city; and the Metropolitan Area of Morelia [3], a rapidly expanding metropolis where population doubled over the last 30 years and a subsidence-creep-fault process has been identified. Surface faulting hazard resulting from differential settlement is constrained via the estimation of angular distortions that are produced on urban structures using the InSAR-derived vertical deformation field, as well as the computation of horizontal strain (tensile and compressive) based on the InSAR-derived E-W deformation field. A methodology to embed such information into the process of risk assessment for urban infrastructure is proposed and demonstrated. The results and discussion will showcase how the InSAR deformation datasets and their derived products are essential not only to constrain the deformation processes, but also to provide a valuable input for the quantification of properties and population at risk. InSAR-derived evidence of land subsidence and its induced risk could be a crucial information source for groundwater management policy makers and regulators, to identify the most impacted cities and optimize groundwater management and development plans to accommodate existing and future water demands.

[1] Cigna F., Tapete D. 2021. Present-day land subsidence rates, surface faulting hazard and risk in Mexico City with 2014-2020 Sentinel-1 IW InSAR. Remote Sensing of Environment, 253, 1-19, https://doi.org/10.1016/j.rse.2020.112161

[2] Cigna F., Tapete D. 2021. Satellite InSAR survey of structurally-controlled land subsidence due to groundwater exploitation in the Aguascalientes Valley, Mexico. Remote Sensing of Environment, 254, 1-23, https://doi.org/10.1016/j.rse.2020.112254

[3] Cigna F., Osmanoǧlu B., Cabral-Cano E., Dixon T.H., Ávila-Olivera J.A., Garduño-Monroy V.H., DeMets C., Wdowinski S., 2012. Monitoring land subsidence and its induced geological hazard with Synthetic Aperture Radar Interferometry: A case study in Morelia, Mexico. Remote Sensing of Environment, 117, 146-161. https://doi.org/10.1016/j.rse.2011.09.005

United Arab Emirates (UAE) characterised by arid climate with limited resources of fresh water and high-water demand in sectors of domestic, agriculture, and industry. Due to this limitation in water resources, sustainable groundwater practices are required to maintain the available resources to diminish. One of the most crucial groundwater practices are monitoring of groundwater dynamics, in quality and quantity, and the implications of unsustainable groundwater usage.
Al Ain region is located at the eastern part of Abu Dhabi Emirate, UAE at the border with Sultanate Oman. This region occupies 50% of the Abu Dhabi Emirate agricultural activities consumes huge amount of groundwater with annual discharge more than 200 million m3. The groundwater resources can be found at the unconfined gravel aquifers and sand dune aquifers.
The current study aims at investigating the deformations occurring at the site due to the overexploitation of the aquifers by combining SAR satellite data, with ground water level measurements and ground truth surveys. Sentinel-1 data, provided by the European Space Agency (ESA), were used to process the SAR interferometry over the study area. Water level data were provided by the Environment Agency of Abu Dhabi (EAD), and they were used to determine zones affected by groundwater overexploitation. Land surfaced subsidence evidences were identified in the field, confirming the deformations identified by the SAR interferometry product. The dataset used consists of 37 Sentinel-1A Single Look Complex (SLC) images acquired along the ascending orbit from path 130 and frames 73 and 75, for a time span between February 2015 and May 2019. The image acquired on 22 October 2017 was selected as a primary, or master, image to increase the expected coherence due to it is minimum spatial and temporal baselines.
The water level dataset indicated an extensive cone of depression covering the area under investigation from 2013 to 2019. It is clear that the extended network of irrigation wells has systematically affected the unconfined sand dune aquifers unit and resulted in lowering the groundwater level with a maximum drawdown at its centre of approximately 40 to 50 m. As expected, at the perimeter of the cone, the ground water lowering gradually decreases in relation to the distance from the centre of the cone. The great discharge from the aquifer, more than 240 million m3, along with a very low hydraulic conductivity of the aquifer resulted in a low annual groundwater recharge.
The study revealed an extensive land surface subsidence with a rate of 40 mm/year in the period between 2015 and 2019. The cone of depression for the water level drawdown in the study area was found in spatial correlation with the detected land surface subsidence bowl. This can be concluded that the land surface subsidence was triggered by the groundwater over extraction.
Furthermore, it was proved that the repeat-pass satellite SAR interferometry can provided substantial information about the actual extent of the land subsidence phenomenon. Space-based technologies are cost effective, providing high spatial coverage. So, they are able to fill the data and knowledge gaps and reduce the uncertainties by providing high spatial and temporal valuable information about the extend and the progress of the subsidence.

This work was supported by a grant from the United Arab Emirates University (UAEU) National Center for Water and Energy under grant number 31R155-Research Center-NWC-3-2017.

Background
The Netherlands has been actively managing its groundwater table for centuries, and much of the Dutch agricultural sector is based in drained peatlands called polders. The polders are separated into parcels which are rectangular plots surrounded by drainage ditches. It has been observed that groundwater levels are the most significant driver of soil surface height variation in the region [1]. It has been hypothesized that groundwater management regimes in these regions are causing them to irreversibly subside at a rate which is faster than sea level rise [2, 3].

Problem Statement
Direct observation of surface motion of this region has so far not been possible with distributed scatterer (DS) InSAR due to high levels of noise in these grass-covered fields and rapid deformation between consecutive SAR acquisitions [4]. These rapid shifts can frequently result in phase unwrapping errors when typical DS processing techniques are used. Additionally, relating the observed InSAR time series with respect to an absolute reference frame has not been possible due to the decoupling of point scatterer (PS) and DS processing techniques, and the lack of a well-defined reference benchmark. To understand the scope of the effects of subsidence in these regions, and subsequently act on it, scientists and policymakers need to know the subsidence rates in absolute terms, so that they may be compared to other locations, or other hazards such as sea level rise.

Our Approach
We adopt a novel multilooking strategy which uses parcels as the basic unit of measure for a DS InSAR monitoring system. This is a natural choice, as the land cover and water table are almost always consistent over a given parcel. Our approach has several advantages over traditional multilooking strategies, as it simultaneously groups pixels together which are physically acted upon by the same mechanism, while also providing a many-fold increase in the effective number of looks. This significantly reduces measurement noise and ensures that
all the pixels being multilooked are tied to the same ergodic process, which a typical boxcar filter will not do, even when coupled with a statistical homogeneity test.

These multilooked phases can now be treated as virtual point scatterers; they are assigned a location corresponding to the centroid of the parcel in question, and imported into the Delft Persistent Scatterer Interferometry (DePSI) system for time-series analysis as 2nd order points [5]. This allows us to take advantage of the robust point scatterers (PS) in the image to form a 1st order network of points for atmospheric phase screen (APS) removal, trend removal and variance component estimation. This mixed system allows for the simultaneous monitoring of both agricultural zones and the built environment.

Via the connection to the 1st order network, we are able to make an arc connection to the Integrated Geodetic Reference Station (IGRS) [6] in the region. These stations are comprised of a corner reflector, a GPS receiver and other geodetic instrumentation which will allow us to make a direct connection from the observed parcel movement to an absolute geodetic reference frame.

In our contribution we will present the first results of parcel-multilooked DS estimation performed within this mixed scatterer framework of the region surrounding Zegveld, The Netherlands based on Sentinel-1 data. Following the parameter estimation in DePSI, we select an IGRS in the region of interest as the InSAR reference point, and transform the relative displacement time series into an absolute frame using the co-located GPS measurements.

References
[1] S. van Asselen, G. Erkens, and F. de Graaf, “Monitoring shallow subsidence in cultivated peatlands,” Proceedings of the International Association of Hydrological Sciences, vol. 382, pp. 189–194, 2020.

[2] G. Erkens, M. J. van der Meulen, and H. Middelkoop, “Double trouble: subsidence and co2 respiration due to 1,000 years of dutch coastal peatlands cultivation,” Hydrogeology Journal, vol. 24, no. 3, pp. 551–568, 2016.

[3] T. Hoogland, J. van den Akker, and D. Brus, “Modeling the subsidence of peat soils in the dutch coastal area,” Geoderma, vol. 171-172, pp. 92 – 97, 2012. Entering the Digital Era: Special Issue of Pedometrics 2009, Beijing.

[4] Y. Morishita and R. F. Hanssen, “Deformation parameter estimation in low coherence areas using a multisatellite insar approach,” IEEE Transactions on Geoscience and Remote Sensing, vol. 53, no. 8, pp. 4275–4283, 2015.

[5] F. van Leijen, Persistent Scatterer Interferometry based on geodetic estimation theory. PhD thesis, TU Delft, 2014.

[6] R. F. Hanssen, “A radar retroreflector device and a method of preparing a radar retroreflector device", U.S. Patent No. 2018236215, 2018.

Estimating the deformation of the Dutch countryside with satellite radar interferometry (InSAR) is notoriously difficult due to rapid soil movement and low coherence [1]. Previous research [2][3][4] has shown that variations in soil moisture can create significant contributions to the observed interferometric phase due to changes in the dielectric properties in the scattering medium. This exacerbates the problem of correctly tracking the deformation of these soft soils, as soil moisture variations can be caused by changes in the ground water level, which is a primary driver of shallow ground deformation in the region [5]. Quantifying and removing the soil moisture phase contribution before phase unwrapping can therefore be a very helpful step to reduce noise in deformation time series.

Employing the phase closure analysis is a very promising approach to estimate the contribution of soil moisture variations without the need for prior phase unwrapping. A set of three SAR images are interfered with each other circularly to form three multilooked interferograms. The summation of the estimated expected values of the three interferometric phases are called the closure phases [6]. Theory states that these closure phases must equal zero for a point scatterer, however, this statement loses its validity when multilooking is employed to form a series of phases over distributed scatterers. These non-zero closure phases have been shown to be caused in part by geophysical processes (i.e. soil moisture variations), and provide us with an opportunity to mitigate geophysical contributions to the wrapped phases [7]. Phase noise as a result of lack of interferometric coherence also contributes to the phase closure, with the added difficulty that large geophysical phase closures go hand-in-hand with low coherences, which makes the phase closure an inherently noisy observable. Therefore, a multilooking strategy which simultaneously suppresses noise but preserves the geophysical closure phases is required. This is accomplished by averaging a large number of pixels over a given region in which ergodicity is assumed.

The vast majority of the Dutch countryside is used for agriculture and is segmented into rectangular parcels surrounded by drainage ditches. With few exceptions, the land cover and groundwater table within a parcel are consistent. This geographical feature provides us with a natural way to segment the scene into multilooked regions which can be described by a few known parameters such as soil type, land cover and groundwater level. The multilooked phases of the distributed scatterers (DS) are condensed to a representative point measurement, which is imported into the Delft Persistent Scatterer Interferometry (DePSI) system [8]. This allows us to accurately estimate and remove atmospheric phase contributions to the measurement and compare the movement of unstable DS points with nearby high-quality point scatterers (PS) in the region.

Closure phases are formed circularly with consecutive 6-day acquisitions from the parcel-based multilooked interferograms. Based on the assumption that noise in closure phases is Gaussian distributed, a numerical model with the inputs of the corresponding coherences of each interferogram and the number of looks is developed in order to perform a significance test. The model output is the standard deviation of the closure phases if solely induced by decorrelation noise. Significance ratios are then computed by dividing closure phases with the numerically simulated standard deviation to estimate the SNR. We find that 1) the closure phases significantly exceed the noise estimate, which suggests a deterministic cause, and 2) there is a promising correlation between the significance ratio and the variation of in-situ groundwater level measurements, which can be used as a proxy for soil moisture variations [9].

This is the first step towards our overall goal of using the observed closure phase to aid in time series analysis. Estimation of soil moisture variation induced interferometric phase differences can subsequently be used to create a phase screen to remove the soil moisture variation component in the observed InSAR phase prior to unwrapping, for instance by using the preliminary interferometric soil moisture model developed by De Zan et al. in [3]. Based on a quantitative correlation analysis between the closure phase observation and theory, we propose recommendations for removing the unwanted soil moisture variation induced phase from an InSAR time series in order to reduce decorrelation and aid in phase unwrapping. This research improves our understanding of the effects of soil moisture variations on the wrapped interferometric phases, and paves the way for deriving a more accurate deformation time series.

References
[1] Y. Morishita and R. F. Hanssen. Temporal decorrelation in l-, c-, and x-band satellite radar interferometry for pasture on drained peat soils. IEEE Transactions on Geoscience and Remote Sensing, 53(2):1096–1104, 2015.
[2] Andrew K Gabriel, Richard M Goldstein, and Howard A Zebker. Mapping small elevation changes over large areas: Differential radar interferometry. Journal of Geophysical Research: Solid Earth, 94(B7):9183–9191, 1989.
[3] Francesco De Zan, Alessandro Parizzi, Pau Prats-Iraola, and Paco López-Dekker. A sar interferometric model for soil moisture. IEEE Transactions on Geoscience and Remote Sensing, 52(1):418–425, 2013.
[4] Simon Zwieback, Scott Hensley, and Irena Hajnsek. Assessment of soil moisture effects on l-band radar interferometry. Remote Sensing of Environment, 164:77–89, 2015.
[5] S. van Asselen, G. Erkens, and F. de Graaf. Monitoring shallow subsidence in cultivated peatlands. Proceedings of the International Association of Hydrological Sciences, 382:189–194, 2020.
[6] Francesco De Zan, Alessandro Parizzi, and Pau Prats. A proposal for a sar interferometric model of soil moisture. In 2012 IEEE International Geoscience and Remote Sensing Symposium, pages 630–633. IEEE, 2012.
[7] Francesco De Zan, Mariantonietta Zonno, and Paco Lopez-Dekker. Phase inconsistencies and multiple scattering in sar interferometry. IEEE Transactions on Geoscience and Remote Sensing, 53(12):6608–6616, 2015.
[8] F. van Leijen. Persistent Scatterer Interferometry Based on Geodetic Estimation Theory. PhD thesis, TU Delft, 2014.
[9] AH de Nijs and RAM de Jeu. Evaluatie van hoge resolutie satelliet bodemvochtproducten met behulp van grondwaterstandmetingen. Stromingen: vakblad voor hydrologen, 28:23–33, 2017.

On July 23, 1972, the Earth Resources Technology Satellite (ERTS-1), later renamed as Landsat, was launched into orbit. This was the first civil satellite doing Earth Observation over our planet Earth. Coincidentally, also in 1972, the World Heritage Convention was agreed upon at the United Nations Educational Scientific and Cultural Organization (UNESCO). The 1972 World Heritage Convention concerning the Protection of the World Cultural and Natural Heritage developed from the merging of two separate movements: the first focusing on the preservation of cultural sites, and the other dealing with the conservation of nature. The most significant feature of the 1972 World Heritage Convention is that it links together in a single document the concepts of nature conservation and the preservation of cultural properties. The Convention recognizes the way in which people interact with nature, and the fundamental need to preserve the balance between the two.
Since the end of the 1990’s, the concept of the potential of using Earth Observation data and technologies to support Natural and Cultural heritage sites has been on the table of discussions.
This paper will present the current state-of-the-art as far as Earth Observation supporting Natural and Cultural heritage. The main concepts described in this paper emerge from the overall experience gained during the fifteen years of managing and implementing the European Space Agency (ESA) and UNESCO Open initiative on the use of space technologies to support World Heritage sites: “From Space to Place”, as well as a series of further additional activities supporting heritage sites implemented afterwards.
Although the World Heritage Convention deals with Natural and Cultural heritage under the same framework Convention, the issue of EO supporting Natural and Cultural Heritage has to be addressed as dealing with two different concepts of Heritage, therefore through a separated approach. The needs and requirements of Natural heritage are different to those of the Cultural heritage domain. The main know how to manage Natural heritage is basically under the “hat” of for example, a park ranger, while the expertise to manage a Cultural heritage site, depending on the type of cultural heritage site, might be for example under the “hat” of an archaeologist. This implies per se that the chain of activities for the management, conservation, monitoring and dissemination of a Natural heritage site is different to the “chain” of activities required for a Cultural heritage site. Therefore, the EO services required to support Natural heritage sites are different to those required to support Cultural heritage sites. However, we have identified that there is some overlap, e.g. some few services can be used for both Natural and Cultural heritage sites.
An innovative technological emerging issue in the area of Cultural heritage is that the state-of-the-art enables the development of associated Digital Twins, this means the elaboration of a digital virtual representation of the Cultural heritage object (or site) that serves as the real-time digital counterpart of the physical Cultural Heritage object. A digital virtual tour of some heritage sites is starting to be available, further enhanced with virtual reality and virtual augmented reality. This is an area where there might be some overlap between the technologies required to implement ESA’s Digital Twin Earth concept, and the current technologies being used to develop digital twin cultural heritage objects.
The presentation will cover the current status for Natural heritage sites, this type of heritage is basically related to preserving selected Earth ecosystems. Since most of the EO satellites have been designed to monitor the Earth ecosystems, then, in the area of natural heritage sites, there are already some EO operational services being used for the benefit of Natural heritage. Major emphasis will then be done on the complexity related with Cultural heritage sites, the main problems identified, the current state-of-the art of research as well as some ideas to encourage discussion on what can be done to start working jointly towards the development of associated EO services for the benefit of Cultural heritage.
An eventual market for EO-based services to support Natural heritage and a potential market to support Cultural heritage might be in the process of building up. Some suggestions on how we could further encourage this momentum will be addressed.

Earth Observation and ground remote sensing non-contact technologies are considered innovative methods to support decision making and site management sustainable exploitation of cultural assets. The broad spectra of remote sensing techniques provide the ideal platform to undertake a wide range of practical, cost-efficient, and easily programmable studies, not easily acquired with other tools. In the case of cultural landscapes, these techniques offer the opportunity to ensure repeated monitoring of multiple parameters in a macro and micro spatial scale, offering European broad comparisons and contrasts.

Over the past few years, the advancement of satellite observations and space-based products and the availability of open-source software have revolutionised archaeological practices. As indicated by the relevant literature, Earth observation sensors have been widely adopted for archaeological purposes in the recent past [1-2]. In parallel, artificial intelligence image-based methods have also been populated in the relevant literature [3]. The contribution of the European Copernicus and other international space programs that provide free and full access to a range of datasets has been instrumental for the broader use of space-based observations for heritage monitoring [4-5]. Nevertheless, local stakeholders and policymakers call for concrete and tangible outcomes for their use [6].

This presentation summarises recent applications based on the European Copernicus space program and other international space programs for cultural heritage applications in Cyprus, developed through two collaborative research projects. The results presented here are part of NAVIGATOR [7] (Copernicus Earth Observation Big Data for Cultural Heritage, EXCELLENCE/0918/0052) and PERIsCOPE [8] (Portal for heritage buildings integration into the contemporary built environment, INTEGRATED/0918/0034).

Under the NAVIGATOR project, an overview of the Earth Observation contribution to cultural heritage disaster risk management is discussed. The overall risk cycle and the potential links between the space-based technologies for detection, monitoring and analysis of cultural heritage sites are presented [9]. In addition, the use of optical and radar Sentinel missions for detecting displacements within archaeological sites in Cyprus after a 5.6 magnitude scale earthquake event through big data cloud platforms such as the Hybrid Pluggable Processing Pipeline (HyP3) system are discussed (Fig. 1, [10]).

Change detection methods using optical images are presented, supporting thus needs for landscape studies and mapping the broader context of an area [11]. A case study for detecting fire intensity through optical Sentinel-2 images are also presented, implemented in a recent fire event in Cyprus (Aug. 2021). Under the PERIsCOPE project, the use of thermal Landsat-7 and -8 images for detecting hot-spot areas in the municipality of Limassol and Strovolos are discussed, supported by Google Earth big data cloud platform (Fig.2, [12]). In addition, the Google Earth platform was used to extract vegetation properties from Landsat 8 [13-14] and identified changes in vegetation cover over the two areas (Limassol and Strovolos) (Figure 3).


These examples are considered indicative concerning the broader aspects of how Copernicus and other space programs can support real needs of heritage protection and management. Their use and further elaboration can only benefit from integrating and adopting best practices by responsible stakeholders and policymakers.

Acknowledgements

The authors would like to acknowledge the NAVIGATOR project, co-funded by the Republic of Cyprus and the European Union's Structural Funds in Cyprus under the Research and Innovation Foundation grant agreement EXCELLENCE/0918/0052 (Copernicus Earth Observation Big Data for Cultural Heritage). Results related to the Landsat thermal analysis are part of the "Portal for heritage buildings integration into the contemporary built environment", in short PERISCOPE, co-financed by the European Regional Development Fund and the Republic of Cyprus through the Research & Innovation Foundation. Grant Agreement INTEGRATED/0918/0034. Lastly, the authors would like to acknowledge the project “Programma Operativo Nazionale Ricerca e Innovazione 2014-2020 - Fondo Sociale Europeo, Azione I.2 “Attrazione e Mobilità Internazionale dei Ricercatori” – Avviso D.D. n 407 del 27/02/2018” CUP: D94I18000220007 – cod. AIM1895471 – 2.

Figure Captions

Figure 1. (a) Unwrapped interferogram. (b) Vertical displacements. (c) Coherence map, enveloping important archaeological sites of Cyprus (Nea Paphos, Tombs of the Kings and the historic town centre) (source [10]).

Figure 2. Seasonal mean temperature over the Strovolos area (Cyprus), between the years 2013 and 2020. The red colour indicates higher mean temperatures, while the blue colour indicates lower mean temperatures.

Figure 3. NDBaI2 index estimated on Strovolos (Cyprus) (a) and Limassol (Cyprus) (b) over the period 2013 (on the left) - 2010 (on the right). NDBaI2 value is comprised between -1 (blue) and 1 (white).

References

[1] Agapiou, A.; Lysandrou, V. Remote Sensing Archaeology: Tracking and mapping evolution in scientific literature from 1999–2015. J. Archaeol. Sci. Rep. 2015, 4, 192–200.
[2] Luo, L.; Wang, X.; Guo, H.; Lasaponara, R.; Zong, X.; Masini, N.; Wang, G.; Shi, P.; Khatteli, H.; Chen, F.; et al. Airborne and spaceborne remote sensing for archaeological and cultural heritage applications: A review of the century (1907–2017). Remote Sens. Environ. 2019, 232, 111280.
[3] Orengo, A.H.; Conesa, C.F.; Garcia-Molsosa, A.; Lobo, A.; Green, S.A.; Madella, M.; Petrie, A.C. Automated detection of archaeological mounds using machine-learning classification of multisensor and multitemporal satellite data. Proc. Natl. Acad. Sci. USA 2020, 117, 18240–18250.
[4] Tapete, D.; Cigna, F. Appraisal of Opportunities and Perspectives for the Systematic Condition Assessment of Heritage Sites with Copernicus Sentinel-2 High-Resolution Multispectral Imagery. Remote Sens. 2018, 10, 561
[5] Zanni, S.; De Rosa, A. Remote Sensing Analyses on Sentinel-2 Images: Looking for Roman Roads in Srem Region (Serbia). Geosciences 2019, 9, 25
[6] Rączkowski, W. Power and/or Penury of Visualizations: Some Thoughts on Remote Sensing Data and Products in Archaeology. Remote Sens. 2020, 12, 2996. https://doi.org/10.3390/rs12182996
[7] Copernicus Earth Observation Big Data for Cultural Heritage, http://web.cut.ac.cy/navigator/ (accessed on 23th Nov. 2021)
[8] Portal for heritage buildings integration into the contemporary built environment, https://uperiscope.cyi.ac.cy (accessed on 23th Nov. 2021)
[9] Agapiou A., Lysandrou V. Hadjimitsis D.G. Earth Observation Contribution to Cultural Heritage Disaster Risk Management: Case Study of Eastern Mediterranean Open-Air Archaeological Monuments and Sites. Remote Sens. 2020, 12, 1330, https://www.mdpi.com/2072-4292/12/8/1330
[10] Agapiou A., Lysandrou V. Detecting displacements within archaeological sites in Cyprus after a 5.6 magnitude scale earthquake event through the Hybrid Pluggable Processing Pipeline (HyP3) cloud-based system and Sentinel-1 Interferometric Synthetic Aperture Radar (InSAR) analysis, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2020, 13, 6115-6123.
[11] Agapiou A., UNESCO World Heritage properties in changing and dynamic environments: change detection methods using optical and radar satellite data. Heritage Science 2021, 9, 64. https://doi.org/10.1186/s40494-021-00542-z.
[12] Agapiou, A.; Lysandrou, V. Observing Thermal Conditions of Historic Buildings through Earth Observation Data and Big Data Engine. Sensors 2021, 21, 4557. https://doi.org/10.3390/s21134557
[13] Capolupo, A., Monterisi, C., Caporusso, G., & Tarantino, E., Extracting Land Cover Data Using GEE: A Review of the Classification Indices. In International Conference on Computational Science and Its Applications, 2020, 782-796, Springer, Cham.
[14] Capolupo, A., Monterisi, C., Saponaro, M., & Tarantino, E. (2020, August). Multi-temporal analysis of land cover changes using Landsat data through Google Earth Engine platform. In Eighth International Conference on Remote Sensing and Geoinformation of the Environment (RSCy2020), 11524, p. 1152419, International Society for Optics and Photonics.

This paper presents an assessment of the use of the DESIS sensor, the imaging spectrometer mounted on the International Space Station (ISS), for the detection of burned areas in sensitive areas. Each DESIS acquisition records continuous spectral information over areas of 30 km × 30 km, a suitable size for such applications, in the visible and near infrared ranges across 235 spectral bands. As DESIS is the first hyperspectral sensor allowing rapid revisit of any site of interest excluding extreme high latitudes, pre- and post-event images can be available, where burned areas can be detected with change detection techniques coupled with suitable, narrow-band spectral indices. Such products may help in timely raising awareness on the endangerment of cultural and natural heritage sites and landscapes, emphasising the importance of Earth Observation (EO) data for monitoring, digitizing and documenting valuable cultural heritage sites.

A first assessment for the case of the Arakapas fire in Cyprus is presented. This event started on Saturday, the 3rd of July 2021 in the Limassol district near the village of Arakapas and was controlled after approximately 24 hours. The area affected by the fire is designated as an area of special aesthetic value of the Troodos mountain range to the South West Shores and is included in the Troodos UNESCO global geo-park, which characterizes it as a natural heritage landscape. According to the Department of Antiquities, there are 13 cultural heritage sites in the extended region of the fire. Indeed, several churches of significant cultural value were in danger, being located close to the fire. DESIS acquisitions in cloud-free conditions are available for the pre- and post-event dates of the 10th of June and 31st of July 2021, respectively. The difference of the narrow-band Normalized Differential Vegetation Index (NDVI), using the narrow bands centered around 620 and 700 nm respectively was used to identify the burned area. Results are favourably matched to available coordinates of known burned sites, and the affected area looks overall well identified according to the available information on the event. Short wave infrared (SWIR) information is usually characterized by relevant emissions in presence of fires and widely used for this kind of analysis. Nevertheless, results show that DESIS data yield precise burnt area maps, in spite of the lack of this spectral information.

Also 10 spectral bands of multispectral Sentinel-2 images from the 12th of June and 27th of July, with spatial resolution between 10 m and 20 m and a swath width of 290 km, were used to calculate different indices frequently applied for burned area assessment using EO data, such as the Normalised Burn Ratio (NBR), Burned Area Index (BAI), and dNBR (differential NBR) Results from these broadband indices are accurate, and are subsequently compared to the results of the narrowband outcomes from DESIS.

The standardization of scanning the Earth’s surface through multiple devices and sensors, coupled with the increasing availability of open access datasets and the development of algorithms for automatic analyses (especially for image recognition and classification), has allowed Landscape Archaeology to utilize such technologies for conducting valuable new studies. Nowadays, the evaluation of spatial and temporal patterns revealed via analyses of artefacts, sites, landscapes, and cultural phenomena is a fundamental step of every archaeological project. This approach facilitates comparisons and correlations between different types of information in relational databases and GIS, enriching the information derived from ground-level fieldwork and lab processing, improving the accuracy of the results and – most importantly – opening new lines of investigation.
The study presented here integrated archaeological and digital methodologies for the identification and classification of different archaeological features related to the long-term nomadic and semi-sedentary behavior of the hunter-gatherer and early pastoral groups occupying the Egyptian Western Desert (hereafter EWD) during the Early and Mid-Holocene.
In this period, the region played a key role as a zone of transition, transformation and exchange between the Sahara and the Nile Valley. These human groups living in the EWD experienced major environmental changes and repeated climatic oscillations that triggered transformations in their economy, mobility, and settlement patterns. Between the 7th and 6th millennia BC, the area of Wadi el Obeiyid in the Farafra Oasis witnessed the presence of semi-sedentary settlements, possibly related to a phase of increasing demographic pressure. The groups living in the area exploited the environment through a mixed economy based on hunting activities, gathering wild plants, ostrich exploitation, and caprine herding; they moved across long distances, especially for procurement of raw materials.
This study focused on two kinds of architectural features that should be related to these groups: slab structure sites, and a specific type of surface fireplace, named Steinplatz. The slab structures are features made of stone slabs vertically stuck in the silt layers and arranged in circular or oval shapes. They are believed to be foundations that originally supported perishable materials forming hut- or shelter-type structures. The Steinplatz hearths are burnt and fire-cracked stone-filled pits and are ubiquitous in the Sahara. They provide evidence of short-term encampments of people moving within the region and are an important archaeological marker of changes in patterns of the past human presence that this work aims to assess. Both slab-structures and Steinplatz have also been identified in other areas of the EWD (Dakhla, Kharga, Great Sand Sea, Gilf Kebir, Karkur Talh, and Jebel Uweinat).
Despite the efforts of several international research groups, due to the difficulty of ground-based surveys, especially after 2015, when access to the EWD was denied by the Egyptian Authorities due to security and safety concerns, identifying and assessing the distribution of these features is essentially impossible via traditional ground-based surveys. The application of a remote, automated, precise, and cost-effective method for feature detection will help us to overcome these obstacles and enrich the existing datasets.
The process applied in this project involved the application of appropriate (established) machine learning algorithms for image recognition (coded in Python/JavaScript) on multispectral and multitemporal satellite imagery, at very high and high resolution (0,6 to 10 m for optical imagery and 1 to 10 m for radar data), and different modalities of acquisition. The imagery derives from different sources (Hexagon KH-9, QuickBird-2, Sentinel 1 and 2, COSMO-SkyMed), and was assembled and processed through the Google Earth Engine Cloud Computing Geospatial Platform (hereafter GEE). The algorithms were coded to factor-in many relevant characteristics and archaeological proxy indicators (crop-, soil-, dump- and shadow-marks; archaeological and environmental hotspots; elevation and microtopography; surface roughness and discontinuity; weather conditions, electric conductivity and moisture content percentage of the soil). The training of the algorithms was performed on vectorized verified datasets of overlapped composite images (Multi-Temporal Aggregates) containing features already known in the archaeological record. The trained algorithms were then applied to an unverified satellite dataset of the same sources. The new results were collected, vectorized and checked within a relational database and a GIS through Error Rate with Cross-Entropy Validation.
The results of this study will provide essential information about human mobility patterns between the Eastern Sahara and the Nile Valley during the Holocene and will shed new light on the contribution of the Saharan communities in the emergence of the Egyptian civilization.

Archaeological and cultural heritage are precious and fragile assets that need to be preserved from degradation and, at the same time, need a proper valorization allowing people to easily access their historical, archaeological and material consistence. The methodological approach to the Cultural Heritage assets protection must therefore be multi-technology, multi-scale and multi sensor, in order to collect data from different data sources, whose correlation will allow a deeper comprehension of the phenomena in act and their best prevention.
Pomerium is a project under development proposed by an Italian consortium in the framework of ESA call “5G for l’ART” and aimed to demonstrate the applicability and effectiveness of a stack of multi-technology instruments and methodologies for the monitoring of the CH exposed to the multiple aggressions of the urban environment. The Area of Interest of the project is the historical center of Rome, with particular reference to Colosseum, Cestio Pyramide, Aurelian Walls and the urban path of Tiber.
Reference Users of the project, involved since the beginning in the planning and operational activities, are:
- Soprintendenza Speciale Archeologia, Belle Arti e Paesaggio of Rome (MIC);
- Parco Archeologico of Colosseum (MIC);
- Sovrintendenza Capitolina (Municipality of Rome).
The project articulates around four main use cases referred to the main risks that environmental and anthropic factors represent for CH:
- Ground or structure instability;
- Pollution;
- Waste and non-authorized soil use;
- Weed vegetation.
For each phenomena to keep under monitoring, a different set of technologies was combined in different configuration, in order to achieve a deep level of knowledge on the active phenomena and on the status of the exposed CH assets.
In detail, the scopes and methodologies of the different use cases are:
Ground or infrastructure instability:
Monitor the displacement phenomena affecting the interested CH assets and their environment, in order to identify and prevent damages and losses through the usage of the following instruments:
- DIFSAR interferometry from Cosmo SKY-Med data for the monitoring of displacements over time;
- On site-displacement sensors applied to a set of points identified by the remote analysis as most affected by displacement phenomena.
Pollution:
Monitor the distribution of urban pollutants around the CH assets and foresee the degradation effect on exposed surfaces through the use of:
- On site air quality sensors integrated by public networks data (ARPA and Municipality of Rome);
- Spray dispersion model, to simulate the distribution of pollutants around the interested surfaces;
- Regression model, to foresee the entity of regression phenomena on marble surfaces due to pollutant agents.
Waste, soil use:.
Detect illicit uses of the soil in protected environments, detect and monitor the presence of waste and debris. The monitoring activities interest the Tiber stretch in urban area (from Flaminio to Marconi) and count on:
- RPAS surveys with optical and IR camera and 360 camera;
- Interactive Virtual Reality environment for the representation of the monitoring results and as an operating environment in which the User may navigate the monitored areas in a realistic mode and identify features of his own interest.
Weed vegetation:
Detect and monitor the growth of weed vegetation on historical surfaces through the use of:
- RPAS surveys with optical and IR camera
- Interactive Virtual Reality environment.
The core of Pomerium is represented by the Web-GIS platform AWARE that will act as unique point of access for the Users to data and analysis produced in the different project’s scenarios. Users will be allowed to data consultation, data extraction and production of thematic reports useful to support their ordinary and extraordinary conservation tasks.

Climate change presents new challenges to ecosystems worldwide. Many ecosystems are currently in a state of transformation due to changing site factors, including the increase in annual mean temperature and the reduction in summer precipitation maxima combined with an increase in winter precipitation. These changes are particularly evident in the orchard meadow ecosystem. Orchard meadows provide habitats for numerous animals and plants which are considered extremely species-rich. Since they are extensively managed landscape elements, orchard meadows are not only affected by the prevailing site factors, they are also dependent on anthropogenic management and cultivation.
In southwestern Germany orchard meadows represent a precious cultural landscape element, where the domestic fruit tree stands are dominated by apple, pear, walnut and plum species. However, a dramatic decline in orchard tree populations was observed in the last decades, which can be attributed on the one hand to climate change effects and on the other hand to deficient management and cultivation. In this context, the primary objective of the project within this study will be to conserve orchard meadows and their multiple benefits in terms of ecosystem services. Furthermore, the objective will be to apply and adapt innovative approaches as well as to improve existing remote sensing techniques.
Forecasts indicate that climate change will increase the occurrence and severity of droughts and dry periods in Baden-Württemberg in the future. This fact confronts orchard meadow stands with numerous serious challenges. Already today, immense damage, which is regarded as the result of droughts, was documented at many sites. Droughts were one of the major causes for the mortality of orchard trees and for an increase in their sensitivity to illnesses and insect damage. Considering this challenge, the present study focuses on the detection, monitoring and evaluation of possible effects of the drought since 2016 on orchard stands and fruit trees. Thereby the study site covers an area of 11ha (about 1,100 trees) at the bottom of the Swabian Alb, which is representative for the local cultural landscape in regards of climatic and geomorphological conditions.
The first part of the approach detects single trees using nDSM and NDVI thresholding of the UAV data. The following part evaluates the development of the NDVI values for the identified tree pixels on the basis of Planet and Sentinel-2 data.
The underlying data are based on UAV flights with a 10-band multispectral camera. In the subsequent analysis of the data DSM and orthomosaics were generated by Structure-from-motion (SfM) techniques. Single trees in the high-resolution aerial images were detected through threshold method based on normalized digital surface models (nDSM) and the normalized difference vegetation index (NDVI). Objective of this process is to detect the tree crown footprints of the individual orchard trees. To analyze drought effects on orchard trees, the vegetation index NDVI will be calculated to estimate tree vitality based on Planet imagery time series as well as Sentinel-2 data. Planet imagery provides a suitable spatial (3 meter) and temporal resolution for crown detection of single orchard trees. Moreover, with regard to data consistency, data availability and radiometric quality the suitability of Sentinel-2 imagery will be tested and evaluated to estimate drought effects. In addition, a statistical evaluation in high-interval time series for each crown footprint was conducted over the study period (2016-2021).
It can be expected that the absent or lower water availability of the trees during the drought periods will be recognized as a deviation in the NDVI pixel trajectories of the single orchard trees. In combination with climate data from the surrounding German Weather Service (DWD) climate stations, the NDVI could serve as indicator for drought stress in orchard meadow stands.
This methodology intends to describe and evaluate the impact of drought on orchards and fruit production. Because these methods are transferable to other areas, the climatic effects on different conditions could be evaluated. Therefore, this method allows to determine future favourable but also unfavourable sites for fruit growing, especially for orchards meadows.

Swarm is an ESA Earth Explorer Mission, launched in November 2013, dedicated to unravel one of the most fundamental aspects of our planet: The Earths magnetic field from the Core to the Magnetoshpere. The Swarm mission consists of three identical satellites (Swarm-Alpha, -Bravo and -Charlie) placed in two different polar orbits - the two lower spacecrafts side-by-side at approximately 440 km and the upper at approximately 510 km. In addition the Third-Party e-POP mission adds the fourth spacecraft to the Swarm constellation as Swarm Echo

This presentation will provide an overall status of the Swarm mission and products and an outlook for the short and long term future.

Swarm is the magnetic field mission of the ESA Earth Observation program composed of three satellites flying in a semi-controlled constellation: Swarm-A and Swarm-C flying in pair and Swarm-B at an higher altitude. Its history in-orbit began in the afternoon of the 22nd of Novem-ber 2013, when the three identical spacecraft separated perfectly from the upper stage of the Rockot launcher at an altitude of about 499 km. Control of the trio was immediately taken over by the ESA’s European Space Operations Centre (ESOC) in Darmstadt, Germany. Following the successful completion of the Launch and Early Orbit Phase (LEOP), commissioning was con-cluded in spring 2014 and precious scientific data have been provided since then.

In order to deliver extremely accurate data to advance our understanding of Earth’s magnetic field and its implications, each Swarm satellite carries a magnetic package, composed by an Absolute Scalar Magnetometer (ASM) and a Vector Field Magnetometer (VFM), an Electrical Field Instrument (EFI) and an Accelerometer (ACC). Unfortunately Swarm-C, due to a failure in LEOP and commissioning, does not carry an ASM.

Two daily ground station contacts per spacecraft are needed to support operations and downlink the scientific data stored in the on-board Mass Memory Unit. Only one ground station pass, however, is supervised by an operator, while in general the missions routine operations are automated, thanks to the Mission Control System used at ESOC and the relevant expertise in mission automation gained from the other Earth Explorers, from GOCE to CryoSat-2.

Many activities and campaigns have been performed through the years to improve instrument anomalies, such as changing the EFI operation’s concept to a limited number of daily science orbits and scrubbing operations to counteract image degradation. In particular, the operations on the EFI, implying a consistent workload for the teams, have been automated with the creation of a new interface with the University of Calgary, in charge of producing a set of files with the required configuration and requests, that are ingested in the FOS Mission Planning System.

Similarly, in the last years the ASM instrument has undertaken more and more sessions in Burst Mode, producing data at 250Hz on request of the Instruments team. Also this activity has been recently integrated in the automated operations concept, to offer flexibility to target this mode based on the space environment’s short-term evolution.

On the platform side, a few anomalies happened and were reacted upon very quickly, e.g. the Swarm-A science data downlink anomaly in 2020, that was solved by routing all science data to the housekeeping storage and re-designing part of the ground segment’s processing to handle this change of concept.

The operations were not affected by the COVID-19 pandemic, in the worst periods of which the teams were mostly working remotely and the operations supervised at ESOC by a reduced team: no outage of science nor change of core operations concept was necessary due to the pandemic.

On the orbits side, a major orbital manoeuvring campaign was undertaken in 2019 to change the relative local time of Swarm-A and Swarm-C, such to meet Swarm-B when the orbital planes were at the closest angular location in Summer to Winter, 2021. This particular scenario called “counter-rotating orbits” implied Swarm-B counter-rotating with respect to the lower pair in a similar orbital plane, a very exciting opportunity for science. In addition to this, the separation along-track of Swarm-A and Swarm-C was tuned down to 4 seconds, then 2 seconds and then kept variable up to 40 seconds to provide different scenarios for science data collection, requiring in total more than 20 orbital manoeuvres.

The presentation will describe the Swarm specific ground segment elements of the Flight Operations Segment (FOS) and explain some of the challenging operations performed so far during this almost-9-years-long journey, from payloads operations to resolution of anomalies and the last orbital dance during the “counter-rotating orbits”. This will offer an interesting overview of the mission’s satellite trio, ready to collect science during the upcoming Solar Cycle… and more.

When not operated in their experimental vector mode (see “Self-calibrated absolute vector data produced by the ASM absolute magnetometers on board the Swarm satellites: results, availability and prospect” by Hulot et al.), Absolute Scalar Magnetometers (ASM) on board the Swarm satellites can be operated in a so-called burst mode, allowing the scalar magnetic field to be sampled at 250 Hz with 1 pT(Hz)-1/2 resolution in the [1-100 Hz] frequency range. Originally, this mode was only intended to be run during the commissioning phase for ASM instrument validation purposes. However, as early short burst mode sessions revealed the ability of these data to detect whistlers produced by lightning in the ELF (Extremely Low Frequency) band, the decision was later made by ESA and IPGP to run regular burst mode sessions and start producing a new L1b Swarm data product. Regular one-week sessions are currently run on Swarm Alpha and Bravo every month, usually during different weeks, with occasional overlap. These data are processed within IPGP in the context of Swarm DISC activities and delivered to ESA as a L1b product known as the ASM burst-mode L1b data.

These ASM burst-mode L1b data revealing many whistlers with high scientific value for investigating both lightning events and the ionosphere (see “Whistler in ELF detected from LEO: lightning detection and ionospheric monitoring using Swarm satellites and the future NanoMagSat mission” by Coïsson et al.), it was further decided to systematically analyse these data, to identify and characterize all whistler events. This led to the production of a new Swarm L2 daily product (one daily file per satellite, whenever the satellite is running in burst mode), providing detailed information (such as time of occurrence, location of satellite at time of detection, dispersion, energy) about each whistler detected during the burst mode session. This product is also processed within IPGP in the context of Swarm DISC activities and delivered to ESA as a L2 product known as the Swarm L2 whistler product.

In this presentation, both the burst mode L1b data and the L2 whistler product will be described with the purpose of further encouraging their use by the community. Details about the data/product format, content and availability will be provided, as well as the way the processing chains are designed and operated. The possibility of permanently and systematically producing such magnetic burst mode data and whistler products within an even wider ELF frequency range using miniaturized magnetometers on board nanosatellites, such as envisioned in the context of the NanoMagSat constellation proposed as a Scout ESA NewSpace Science mission, will also be discussed.

The ESA Swarm Mission is the fourth Earth Explorer of the agency and it was launched at the end of November 2013. It is composed of three identical satellites (A, B and C) flying in a constellation, which varied throughout the course of the mission. The main objective of Swarm is the analysis and modelling of the Earth’s magnetic field at an unprecedented accuracy, through combined measurements of scalar and vector magnetometers. In addition, each satellite carries a set of other instruments, which are useful not only to nominal operations, but also to achieve secondary mission’s objectives. In particular, each of them carries an accelerometer intended to measure the non-gravitational forces acting on each single satellite. However, as this instrument did not function as expected, a significant amount of post-processing was necessary to calibrate the signal, often disturbed by spikes, steps and other types of noise. The calibration focused mainly on Swarm C, for which the data are available since the beginning of the mission and are disseminated monthly. Recently, a Swarm A dataset was also released and it comprises some months of the early mission phase, i.e. February - October 2014, when the solar activity was quite high. Swarm A and Swarm C are the so-called “lower pair” because they have been flying for most of the mission side by side, with a variable separation of four to ten seconds. Therefore, the calibrated accelerometer measurements that they provide, when compared, should be in line with the aforementioned separation.

In this work, an analysis on the newly available Swarm A data is presented, in correlation with data of Swarm C for the same period. After applying a high pass filter, very similar signatures are visible, both at the equator and at the poles, for both satellites. These features agree with signals measured by other instruments on board, and they are in line with current literature on LEO satellites (e.g. CHAMP). Having Swarm A and Swarm C data sets available for overlapping period, allows to finally exploit the accelerometer data from the lower pair of the Swarm constellation. For the first time, scientific results are obtained from two Swarm accelerometers, some of which enable Joule heating estimates, to address one of the secondary objectives of the Swarm mission. Therefore, this analysis could be a precursor of a possible full exploitation of the constellation in case Swarm B calibrated accelerometer data will become available.

Current Swarm density estimations rely on key assumptions to infer these ionospheric parameters from ion admittance on the Langmuir Probe. Namely, the existing methodology assumes a 100% O+ plasma, ignores any ion drifts, and neglects any plasma sheath effects. In a realistic plasma environment, these assumptions are expected to be routinely violated, particularly on the nightside (where light ions constitute a significant minority of plasma) and in the auroral zone (where ion flows are non-negligible). This compromises the accuracy of the existing density product, leading at times to significant residuals when compared with independent estimates such as the International Reference Ionosphere or ground radar conjunctions. Here we report on the status of the ESA DISC SLIDEM project, which aims to relax the above assumptions by incorporating current data from the frontal Thermal Ion Imager faceplate, thus delivering an improved and more robust ion density product. In addition, by utilizing the additional source of information, instead of being neglected, along-track ion drift and effective mass may be explicitly derived or estimated. These parameters, valuable in their own right to the wider geophysics community, are also compared with independent estimates in the form of empirical models (IRI-2016 and Weimer 2005 electric field model), conjunctions with other satellites and with ground radar sites, and with data from alternate instruments aboard the same spacecraft. Finally, particle-in-cell simulations are carried out using the Ptetra code to ensure that the assumptions inherent in the mathematical methodology are robust and hold for the Swarm spacecraft geometry under realistic ambient plasma conditions. Uncertainties and potential sources of error are identified and discussed. The SLIDEM project is currently in a mature state, with the output ionospheric parameter estimates having been validated against a range of independent benchmarks. It is therefore hoped the data products derived using SLIDEM will be widely used by the wider geophysics community to obtain accurate and robust estimates of density, plasma flows, and ion composition.

ESA Swarm Earth Explorer mission consists of three identical spacecraft launched in November 2013 in almost polar orbits, with an altitude of approximately 500 kilometers. Swarm spacecraft are providing very accurate measurements of plasma parameters and magnetic field in the Ionosphere F region for 8 years now, from the various dedicated instruments on-board. Since the beginning of the mission, a particular effort from ESA and Swarm community aimed at improving the data quality, addressing various issues or artefacts in the measurements. This presentation, in particular, illustrates the main findings of a project called SPike-trains in ElecTron TempeRAture measured from Swarm Langmuir probEs (SPETTRALE), focussing on plasma parameters measured by Langmuir probes, and in particular on Electron Temperature (Te).
Since early measurements from Swarm, it was evident that the Te measured by Swarm spacecraft shows sometime very large values, not predicted by the models (e.g. IRI 2016, NeQuick), which have unknown origin. SPETTRALE project has demonstrated that a significant fraction of these Te-Spikes appear for specific orientations of Swarm solar panels with respect to the Sun, and they are ordered as continuous lines as a function of the angle with respect to the Sun, and therefore, could be due to instrumental issues or artefacts.
The second phase of the SPETTRALE project has extended the analysis of Te-spikes as a function of Swarm House Keeping parameters relative to potential and currents from solar panels, in order to clarify the origin of these very high values and characterise the main mechanism and the physical nature of Te Spike-trains observed by Swarm. A statistical analysis has been performed on higher resolution (1 Hz) Swarm House Keeping parameters, acquired during a dedicated campaign in Summer 2021 to improve the analysis of SPETTRALE project.
The main findings of this project, together with the next step to implement a new quality Flag for Swarm Te, will be discussed in this presentation.

The Earth Clouds Aerosol and Radiation Explorer (EarthCARE) mission is an ESA/JAXA multi-instrument mission to launch in 2023. The mission consists of a cloud-profiling radar (CPR), a cloud/aerosol lidar (ATLID), a cloud/aerosol imager (MSI), and a three-view broadband radiometer (BBR) covering both LW and SW bands. The mission will deliver a unique and powerful suite of cloud, aerosol and radiation products.

The EarthCARE lidar is called ATLID (ATmospheric Lidar) and is a high-spectral resolution lidar (HSRL). Like ALADIN (the HSRL lidar carried aboard Aeolus) ATLID operates at 355nm . However, ATLID is optimized for cloud and aerosol sensing, while ALADIN's main focus is on the retrieval of winds. Accordingly, ATLID will measure with a much higher resolution compared with ALADIN and possesses a depolarization channel (unlike ALADIN).

ATLID uses a Fabry-Perot etalon based design to distinguished the thermally broadened return from atmospheric molecules from the spectraly narrower return from clouds and aerosols. This allows for the determination of both the extinction profile as well as the extinction-to-backscatter ratio (also known as the lidar-ratio or S). This is in contrast to elastic backscatter lidars (e.g. CALIPSO) which must specify the S in order to derive the extinction profile. In principle, rather simple direct methods for retrieving extinction and S profiles using HSRL attenuated backscatters exist, however, they require high signal-to-noise ratios. In general, the low SNR of space-based lidars compared to their terrestrial counterparts, leads to the need for extensive horizontal smoothing windows.

Averaging intervals on the order of 100km are often defensible for aerosol fields, but certainly not for cloud observations. For clouds, intervals on the order of a very few kilometers or even finer are necessary. Therefore, it is necessary to separate the "strong" (e.g. cloud) and "weak"(e.g. aerosol) regions at high resolution before smoothing the attenuated backscatter returns. In the ATLID processing chain, this is largely accomplished by using the output of the Featuremask processor (A-FM) which uses adaptive thresholds and ideas drawn from image-processing to provide a high-resolution target mask.

After the cloud-screened signals are averaged, then direct techniques for retrieving the extinction and lidar-ratio profiles at low horizontal resolution are employed. To retrieve the cloud properties, a second pass employing a forward-modelling optimal estimation approach (using the low-resolution results as priors) is applied at high horizontal resolution. The processor that accomplishes this is called the ATLID profile processor (A-PRO). In addition to the retrieval of the optical properties, targets are also classified (e.g. water, ice, aerosol(type)) using e.g. extinction, temperature, lidar-ratio and linear-depolarization ratio) by the ATLID target classification procedure (A-TC) which is embedded within the A-PRO processor.

A-FM and A-PRO have been extensively tested using detailed simulated L1 data derived from the application of advanced lidar radiative transfer models and instrument simulators to high-resolution atmospheric model data. Adaptations of A-FM and A-PRO, called AEL-FM and AEL-PRO respectively, have also been created and successfully applied to ALADIN observations. In this presentation, A-FM and A-PRO will be briefly introduced, new developments highlighted, and various illustrative examples drawn from the set of simulated tests scenes presented and discussed. Additionally, example AEL-FM and AEL-PRO results will be presented.

The EarthCARE satellite will host a range of instruments, all optimised for improving our understanding of the interactions between clouds, aerosols and radiation. Over the past few years, it has been shown, with increasing confidence, that the cloud radar and lidar observations made by EarthCARE could have a direct impact on numerical weather prediction (NWP) forecasts if they are used in data assimilation to help initialise forecasts. Also, as global atmospheric models begin to resolve convective-scale features, the use of observations related to clouds becomes increasingly relevant. The relationship between EarthCARE and NWP data can be seen as symbiotic; the monitoring of satellite observations by comparing with the expected state of the atmosphere can be an invaluable tool for detecting and diagnosing problems before data products are released to the wider scientific community.

Previously, the studies at the European Centre for Medium-Range Weather Forecasts (ECMWF) investigating the potential for the direct impact of EarthCARE observations on NWP via data assimilation have been limited to radar reflectivity and lidar backscatter. In the first part of this talk, we will extend the capability to assimilate other EarthCARE observables, including Doppler velocity, Rayleigh backscatter and particulate extinction. The potential information content of each observation type will be discussed with reference to the Jacobians of the observation operators, which provide the sensitivity of the simulated observations to changes in the model state. The second half of the talk will evaluate the synergy between the suite of EarthCARE observations and the rest of the observing system at ECMWF by using A-Train observations. The interplay between active and passive observations in the minimization will be investigated, for example by assessing the impact of assimilating radar reflectivity on the analysis departures of AMSU-A microwave radiances. Finally, we will discuss future directions for maximising the synergistic power of EarthCARE in NWP, including the use of two-moment observation operators to reduce uncertainty and biases in the simulated observations.

Expanding the diversity of EarthCARE observations that can be simulated within the ECMWF model allows the potential for a more comprehensive quality monitoring system; the potential for further direct impacts of EarthCARE observations on forecast skill to be achieved if they were to be assimilated; greater opportunities for synergistic approaches to reduce uncertainty in the observations; and facilitates the use of the additional observations in model evaluation.

Aerosols are a key actor in the Earth system, with relevance for climate, air quality, and biogeochemical cycles. Aerosol concentrations are highly variable in space and time. A key variability is in their vertical distribution, because it influences aerosols life-time and, as a result, surface concentrations, and because it has an impact on aerosol-cloud interactions.

The instrument ATLID (ATmospheric LIDar) of the EarthCARE mission shall determine vertical profiles of cloud and aerosol physical parameters (altitude, optical depth, backscatter ratio and depolarisation ratio) in synergy with other instruments. Operating in the UV range at 355 nm, ATLID provides atmospheric echoes with a vertical resolution of about 100 m from ground to an altitude of 40 km.

On the side of model-oriented aerosol research (IPCC), the value of using a lidar aerosol simulator has been demonstrated, to ensure consistent comparisons between the modeled aerosols and the observed aerosols. In the current study, we make use of the lidar aerosol simulator implemented within the COSPv2 satellite lidar simulator (Bonazzola et al., in preparation) to estimate the total attenuated backscattered signal (ATB) and the scattering ratios (SR) that would be observed at 355 nm by the lidar ATLID overflying the atmosphere predicted by the E3SMv1 climate model. This simulator performs the computations at the same vertical resolution as the ATLID lidar, making use of aerosol optics from the E3SMv1 model as inputs, and assuming that aerosols are uniformly distributed horizontally within each model grid-box. It applies a cloud masking and an aerosol detection threshold, to get the ATB and SR profiles that would be observed above clouds by ATLID with its actual aerosol detection capability.

Our comparison shows that the aerosol distribution simulated at a seasonal timescale is generally in good agreement with observations, with however a discrepancy in the Southern Hemisphere, as the observed SR maximum is not reproduced in simulations there. Comparison between cloud-screened and non cloud-screened computed SRs shows little differences, indicating that the cloud screening by potentially incorrect model clouds does not affect the mean aerosol signal averaged over a season. Consequently, the differences between observed and simulated SR values are not due to sampling errors, and allow to point out some weaknesses in the aerosol representation in models. The use of several wavelengths can give further indication on the nature of the aerosols that need to be improved.

Supercooled and mixed phase clouds play important roles in high latitude radiation budgets. But models have difficulty in realistically simulating supercooled water clouds and mixed phase clouds. These biases in model clouds lead to local and regional biases in radiation which can produce errors in both weather forecasts and climate projections. CALIOP algorithms classify the ice/water phase of vertically resolved cloud layers and have been widely used to study the global distribution of supercooled water clouds, but CALIOP does not have a mixed-phase cloud type. Observations of high latitude mixed phase clouds needed to guide model improvements are limited to a handful of field campaigns and a handful of ground-based sites.

The current CALIOP cloud phase algorithm discriminates ice clouds from liquid clouds using 532 nm attenuated backscatter and depolarization signals. Both ice and liquid clouds produce depolarization of the lidar backscatter signal but in liquid clouds this depolarization results from multiple scattering, and from single scattering in ice clouds. One consequence is the correlations between the magnitudes of depolarization and backscatter signals are different for ice and liquid clouds. These signals can be visualized in a 2-D diagram and cloud thermodynamic phase can be classified using 2-dimensional thresholds. While mixed-phase clouds are common, in the current CALIOP algorithm cloud layers are only identified as either ice or liquid. However, the CALIOP phase algorithm is applied to layers detected at both 1/3km (single shots) and 5 km horizontal resolution. Investigation of the phase of 1/3 km layers detected within 5 km layers finds that both ice and water may be detected within a 5 km layer classified as liquid. We find the fraction of ice and liquid within 5km layers varies in a way that makes physical sense, and that 5 km cloud layers of mixed phase appear most frequently near the boundaries of the Ice and Water sectors of the current CALIOP cloud phase diagram.

Further information can be obtained from the CALIPSO IIR. The forward model used in IIR cloud retrievals is constrained by information from co-located CALIOP profiles. This approach is used to retrieve effective diameters of both liquid drops and ice particles, providing additional characterization of supercooled and mixed phase clouds that is complementary to the CALIOP measurements. This presentation will discuss the identification and characterization of mixed phase clouds using these observations from CALIPSO. These techniques can be applied to future ATLID and MSI observations, so these results are highly relevant to the EarthCARE mission.

Ice clouds play a crucial role in Earth’s radiation budget since they are semi-transparent to solar radiation while they are very effective in trapping thermal radiation from escaping into space. Active remote sensing with radar and lidar from space can help to quantify this process by measuring the spatial and vertical distribution of ice clouds on a global scale. Here, the variational algorithm VarCloud has been a workhorse for recent studies of global ice cloud properties to reconcile measurements from the A-Train satellites CloudSat and CALIPSO. Looking ahead, the upcoming ESA/JAXA satellite mission EarthCARE will give the opportunity to apply this variational framework on a single platform. With the launch of EarthCARE already on the horizon, the consolidation of a validation strategy of retrieved ice cloud properties is now of outermost importance.

Within this context, the German research aircraft HALO was equipped with an EarthCARE-like payload consisting of a high spectral resolution lidar (HSRL) system at 532 nm and a high-power cloud radar at 35 GHz. In case studies, coordinated flights were performed along other airborne platforms carrying instruments at different wavelengths or in situ probes. In combination with additional underpasses of the A-Train satellites, we learned numerous lessons when cloud properties are compared at different scales. Besides these case studies, we applied VarCloud to a month of airborne data acquired over the North Atlantic to test if a statistical comparison with ice cloud properties from the ERA5 reanalysis model is possible.

In this presentation, we want to share our insights into the effect of different spatial resolutions and different instrument sensitivities on retrieved ice cloud properties. By comparing retrieval results between platforms, we can differentiate between spatial resolution and instrument sensitivity effects. Besides the validation with scarce in situ data, the comparison with microwave constrained ERA5 reanalysis allows us to identify regions with potential biases. This enables us to anticipate if retrieved cloud properties can be compared on a statistical basis or only along coordinated flight legs and if conversions are necessary. By sharing our knowledge with the wider community, we hope to foster helpful discussions to consolidate an airborne validation strategy for EarthCARE.

The objective of the EarthCARE satellite mission is to help improve numerical predictions of weather, air quality, and climatic change via application of synergistic retrieval algorithms to observational data from its cloud-profiling radar, backscattering lidar, and passive multi-spectral imager (MSI). EarthCARE’s overarching scientific goal is to retrieve cloud and aerosol properties with enough accuracy that when used to initialize atmospheric radiative transfer (RT) models, simulated top-of-atmosphere (TOA) broadband radiative fluxes, for domains covering approximately 100 km2, “typically” agree with their observationally-based counterparts to within W m-2. “Observed” TOA fluxes derive from radiances measured by EarthCARE’s multi-angle broadband radiometer (BBR). As the latter are not used by retrieval algorithms, comparing them to modelled values, obtained by RT models operating on retrieved quantities, affects a “moderately stringent” verification of the retrievals; “moderately stringent” because BBR radiances consist, in part, of photons that also constitute to MSI radiances. This imperfection aside, this radiative closure verification is a well-defined and cost-effective final stage in EarthCARE’s formal “production model”. Its purpose is to both: i) provide, at all stages of the mission, quantitative feedback to algorithm developers regarding performance of their algorithms; and ii) help users focus, for whatever reason, on retrievals ranging from those that “appear” to have performed well to those that “appear” to have encountered difficulties.

This presentation discusses the essential features of EarthCARE’s radiative closure assessment. It begins with a review of the Scene Construction Algorithm (SCA) that applies a radiance-matching technique to MSI data and expands the ~1 km wide cross-section of retrieved profiles into the across-track direction thereby enabling application of 3D RT models that simulate TOA radiances and fluxes. It is worth noting that the EarthCARE mission appears to be the first Earth science project to apply 3D RT models in an operational sense; all others use 1D models. For continuity with previous missions, however, conventional 1D RT models are also applied to each retrieved column with results averaged over the same assessment domains used by the 3D RT models. The assessment domains measure ~5 km across-track by ~21 km along-track, Simulated TOA quantities are then compared with radiances measured directly by the BBR and fluxes inferred from those radiances by angular direction models that were developed specifically for EarthCARE. Results of these comparisons will be shown for a virtual observational system that consists of synthetic observations computed for surface and atmospheric conditions simulated by the Canadian Numerical Weather Prediction (NWP) model. These test scenes correspond to full EarthCARE “frames” that measure approximately 200 km by 6,000 km. There are three in total and they cover a wide range of cloud, aerosol, and surface conditions. Horizontal grid-spacing is 0.25 km, which includes some geophysical variability below the resolution of most of EarthCARE’s observations.

The Ice Sheet Mass Balance Inter-Comparison Exercise (IMBIE) led by ESA and NASA aims at reconciling estimates of ice sheet mass balance from satellite gravimetry, altimetry and the mass budget method through community efforts. Building on the success of the two previous phases of IMBIE – during which satellite-based estimates of ice sheet mass balance were reconciled within their respective uncertainties and which showed a 6-fold increase in the rate of mass loss during the satellite era – IMBIE has now entered its third phase. The objectives of this new phase of IMBIE, supported by ESA CCI, are to (i) include data from new satellite missions including GRACE-FO and ICESAT-2, (ii) provide annual assessments of ice sheet mass balance, (iii) partition changes into dynamics and surface mass balance processes, (iv) produce regional assessments and (v) examine the remaining biases between the three geodetic techniques, all in order to provide more robust and regular estimates of ice sheet mass balance and their contribution to global mean sea level rise. In this paper, we report on the recent progress of IMBIE-3. Following the last IMBIE update produced for the IPCC’s sixth assessment report (AR6), for which we extended our time-series of mass change of Greenland and Antarctica until the end of 2020, we are now preparing our next annual update, which will cover the year 2021.

Melting of the Greenland and Antarctic ice sheets currently contributes more than one third of global sea-level rise. As Earth’s climate continues to warm throughout the 21st Century, ice loss is expected to increase further, with the potential to cause widespread social and economic disruption. To track the changes that are currently underway in the Polar regions requires detailed and systematic monitoring programmes. Given the vast and inaccessible nature of the ice sheets, this is only feasible from space. One technique that has proved particularly valuable in recent decades is that of satellite radar altimetry. Since the launch of ERS-1 thirty years ago, ESA satellites have provided a continuous record of ice sheet surface elevation change, and with it valuable information relating to the physical processes that drive ice mass imbalance.

Changes in surface elevation are a signature of multiple physical processes that drive ice sheet mass balance. Long term trends in surface elevation can indicate glacier dynamic imbalance, driven for example, by changes in ocean forcing. High frequency tidal and atmospheric pressure-induced oscillations in ice shelf height can be used to identify glacier grounding lines. Localised uplift and subsidence can be an indicator of the passage of water beneath ice sheets, whilst seasonal cycles in elevation can provide information relating to snowfall, surface ablation, and run-off.

With the ever-increasing volume and resolution of satellite topographic data, comes the greater potential to push the boundaries of the physical and glaciological processes that can be observed. Alongside this, comes the possibility to revisit historical instruments, such as ERS-1, ERS-2 and Envisat, to improve the fidelity and useability of these datasets; to exploit complementary sources of information, such as a new generation of super high resolution, meter-scale, Digital Elevation Models; and to employ advanced statistical techniques, to extract increasing information from the data acquired.

In this presentation, we bring together results from multiple ESA-funded studies, spanning ERS-1 through to the latest Sentinel-3 mission, to show how they are leading to fundamental advances in our understanding of ice sheet processes. Specifically, we describe how the achievements of the Cryo-TEMPO study, the Polar+ Surface Mass Balance study, the Polar+ 4D Greenland study, FDR4ALT and the S3 Land STM MPC come together to improve our understanding of the processes driving present day ice sheet mass imbalance. We will describe the progress made to improve the quality and useability of these measurements, from ERS-1 through to Sentinel-3, and present a number of case studies that show how this has delivered new insight into a diverse range of physical processes, including subglacial hydrology, ice sheet meteorology, grounding line mapping and long-term ice imbalance.

Multi-satellite data assessments have provided evidence for a six-fold increase in mass loss of the Antarctica Ice Sheet since 1992 (Shepherd et al., 2018). Driven mainly by an increase in ice discharge in the Amundsen Sea Embayment, these changes are likely to continue in future, and may indicate surpassing of the stability threshold for the West Antarctic Ice Sheet (Arthern and Williams, 2017). However, due to numerous unknown or poorly known parameters entering ice sheet simulations, the dynamic evolution of the Antarctic ice sheet remains the largest uncertainty in global sea-level projections. Remote sensing offers accurate observations of the ice sheet’s current dynamic state, including its mass balance components, crucial to limit projection ensembles to pathways satisfying present-day remote sensing observations.

Here we focus on isolating regional accelerations of mass change caused by ice dynamics and by surface-mass balance (SMB) in the GRACE/GRACE-FO data for 2002-2021. We quantify the ice-dynamic acceleration in Antarctica based on differencing GRACE/GRACE-FO and SMB, as represented by regional climate models and ERA-5 reanalysis data. We show that this indirect method presents an alternative to estimates based on determining dynamic acceleration from remote sensing of the surface-ice velocity, e.g. using InSAR. We find that with regard to the acceleration component, both the direct and indirect methods produce consistent estimates of dynamic acceleration, with similar uncertainties. Furthermore, we show that accelerations and interannual variations in GRACE/GRACE-FO data are largely driven by SMB variations related to largescale atmospheric circulation patterns. While the apparent acceleration of SMB shows large fluctuations depending on the time period considered, we show that the recovered ice dynamic acceleration is a stationary feature.

With the assumption that mass loss will continue at the present rate, we extrapolate the trends inferred from the satellite data to the year 2100. As a sensitivity experiment we in- and exclude the acceleration of dynamic discharge in the extrapolation. We show that a quadratic pathway consistent with today’s satellite observations produces 7.6 ± 2.9 cm of sea-level rise until 2100, in comparison with a linear extrapolation of the rates only of 2.9 ± 0.6 cm. We show that validation of larger projections ensembles with remote sensing observations is crucial to limit the spread of pathways in the numerical simulations, and, thus reduce the uncertainty in the projected sea-level contribution from Antarctica.

Arthern, R. J., and Williams, C. R. (2017). The sensitivity of West Antarctica to the submarine melting feedback. Geophys. Res. Lett. 44, 2352–2359. doi:10.1002/2017GL072514.
Shepherd, A., Ivins, E., Rignot, E., Smith, B., van den Broeke, M., Velicogna, I., et al. (2018). Mass balance of the Antarctic Ice Sheet from 1992 to 2017. Nature 556, 219–222.

The internal temperature is a key parameter for the ice sheet dynamics. The actual temperature profile is a determinant of ice rheology, which controls ice deformation and flow, and sliding over the underlying bedrock. Importantly, the ice flow in turn affects its temperature profile through strain heating, which makes observed temperature profiles a powerful input for ice sheet model validation. Up to now temperature profile was available in few boreholes or from glaciological models. Recently, Macelloni et al. (2016) opened up new opportunities for probing ice temperature from space with the low-frequency passive sensors. Indeed, at L-band frequency, the very low absorption of ice and the low scattering by particles (grain size, bubbles in ice) allows a large penetration in the dry snow and ice (several hundreds of meters). Macelloni et al. (2019) performed the first retrieval of the ice sheet temperature in Antarctica by using the European Space Agency (ESA)’s Soil Moisture and Ocean Salinity (SMOS) L-band observations. They used the minimization of the difference between SMOS brightness temperature and microwave emission model simulations that include ice temperature emulator based on glaciological models. Here, in the framework of the ESA 4DAntarctica and SMOS Extension projects, we propose two main improvements.

First, a new method based on a Bayesian approach have been developed in order to improve the accuracy of the retrieved ice temperature and to provide an uncertainty estimation along the profiles. The Bayesian inference takes as free parameters: ice thickness, surface ice temperature, snow accumulation and geothermal heat flux (GHF). The parameter space investigation is achieved through a Markov Chain Monte Carlo (MCMC) method. Here, the differential evolution adaptive Metropolis (DREAM) algorithm is used for its performances. It runs multiple different Markov chains in parallel and uses a discrete proposal distribution to evolve the sampler to the posterior distribution (Laloy and Vrugt, 2012). For each SMOS brightness temperature observation, 1000 iterations are run on 5 parallel chains. The 2500 first iterations are discarded (aka. burn-in) and only the last 2500 are used for the final ice temperature profile estimation. The posterior probability distribution captures the most likely parameter set (i.e. a surface temperature, snow accumulation and GHF combination), and so, the most likely ice temperature profiles associated to this SMOS observation. It also provides the standard deviation which inform on the temperature uncertainty along the depth.

Moreover, Macelloni et al. (2019) used an ice temperature emulator based on a one-dimensional model (Robin 1955). As the Robin model neglects the horizontal advection, it can only be used in regions with very slow horizontal ice drift and this limits the retrieval to the Antarctic Plateau. In order to extend the analysis over Antarctica, a three-dimensional glaciological model (GRISLI, Quiquet et al., 2018) was used to generate temperature profiles as inputs for the Bayesian approach. In order to speed up the process, an emulator based on a deep neural network (DNN), was developed to reproduce GRISLI temperature field.

Most of the ice in the Antarctic ice sheet drains from the continent to the ocean through fast-flowing ice streams and glaciers. The high velocity of these features are thought to be maintained by the presence of water at the base of the ice sheet, which reduces friction. Subglacial water moving has been linked to transient glacier flow acceleration and enhanced melt at the grounding line. Therefore, the presence, location, and movement of water at the base of the ice sheet are likely significant controls on the mass balance of Antarctica.

The transport of subglacial water from the interior of Antarctica to the grounding line was once thought to be a steady state process. It is now known that subglacial water collects in hydrological sinks, which store and release water in episodic events. These features can be detected and quantified by satellite altimetry by searching for localized elevation change of the ice sheet’s surface. This behaviour is interpreted to be water moving in and out of ‘active’ subglacial lakes.

Quantifying the volume of water involved in these active events can inform on processes otherwise hidden from view, providing valuable information for simulating the subglacial environment. In particular, the period immediately following a lake drainage is dominated by recharge from the subglacial water draining into the lake, thus providing a mean to quantify the subglacial water fluxes. In some cases, subglacial lakes exist in hydrologically connected groups, providing rich information on such fluxes. One such group is the set of four subglacial lakes which exist beneath the Thwaites glacier. These lakes underwent two drainage events in 2013 and in 2017, with a clear period of recharge between the two events. Estimates of subglacial lake recharge rates were extracted and compared against modelled values. These observed rates of recharge were significantly greater than those produced by the modelled output, which implies subglacial melt production under the Thwaites glacier is underestimated.

Given the significance of subglacial water on the behaviour of ice sheets it is important to derive methods that can constrain and validate subglacial melting rates. Direct observations of the subglacial system are impossible due to the thickness of the ice sheet. Estimates of subglacial melt rates are constrained strictly by models, which take into account the impact of geothermal heat flux, vertical dissipation and frictional heat. With few in-situ observations across Antarctica it is currently not possible to validate the results produced from such models. Here, as part of the 4DAntarctic project, we use CryoSat-2 altimetry to produce time-dependent volume time-series for known active subglacial lakes across Antarctica. From these we can extract recharge rates, which act as a lower bound on subglacial melt production. We compare these values against rates of recharge derived by routing modelled subglacial melt across the Antarctic Ice Sheet bed. This provides us with a unique dataset to explore the sub-glacial environment, comparing direct observations of the subglacial network against the theoretical behaviour.

Ice sheets are a key component of the Earth system, impacting on global sea level, ocean circulation and bio-geochemical processes. Significant quantities of liquid water are being produced and transported at the ice sheet surface, base, and beneath its floating sections, and this water is in turn interacting with the ice sheet itself.

Surface meltwater drives ice sheet mass imbalance; for example enhanced melt accounts for 60% of ice loss from Greenland, and while in Antarctica the impacts of meltwater are proportionally much lower, its volume is largely unknown and projected to rise. The presence of surface melt water is also a trigger for ice shelf calving and collapse, for example at the Antarctic Peninsula where rising air and ocean temperatures have preceded numerous major collapse events in recent decades.

Meltwater is generated at the ice sheet base primarily by geothermal heating and friction associated with ice flow, and this feeds a vast network of lakes and rivers creating a unique bio-chemical environment. The presence of melt water between the ice sheet and bedrock also impacts on the flow of ice into the sea leading to regions of fast-flowing ice. Meltwater draining out of the subglacial system at the grounding line generates buoyant plumes that bring warm ocean bottom water into contact with the underside of floating ice shelves, causing them to melt. Meltwater plumes also lead to high nutrient concentrations within the oceans, contributing to vast areas of enhance primary productivity along the Antarctic coast.

Despite the key role that hydrology plays on the ice sheet environment, there is still no global hydrological budget for Antarctica. There is currently a lack of global data on supra- and sub-glacial hydrology, and no systems are in place for continuous monitoring of it or its impact on ice dynamics.

The overall aim of 4DAntarctica is to advance our understanding of the Antarctic Ice Sheet’s supra and sub-glacial hydrology, its evolution, and its role within the broader ice sheet and ocean systems.

We designed our programme of work to address the following specific objectives:

Creating and consolidating an unprecedented dataset composed of ice-sheet wide hydrology and lithospheric products, Earth Observation datasets, and state of the art ice-sheet and hydrology models
Improving our understanding of the physical interaction between electromagnetic radiation, the ice sheet, and liquid water
Developing techniques and algorithms to detect surface and basal melting from satellite observations in conjunction with numerical modelling
Applying these new techniques at local sites and across the continental ice sheet to monitor water dynamics and derive new hydrology datasets
Performing a scientific assessment of Antarctic Ice Sheet hydrology and of its role in the current changes the continent is experiencing
Proposing a future roadmap for enhanced observation of Antarctica’s hydrological cycle

To do so, the project will use a large range of Earth Observation missions (e.g. Sentinel-1, Sentinel-2, SMOS, CryoSat-2, GOCE, TanDEM-X, AMSR2, Landsat, Icesat-2) coupled with ice-sheet and hydrological models. By the end of this project, the programme of work presented here will lead to a dramatically improved quantification of meltwater in Antarctica, an improved understanding of fluxes across the continent and to the ocean, and an improved understanding of the impact of the hydrological cycle on ice sheet’s mass balance, its basal environment, and its vulnerability to climate change.

The Lake Water products within the Copernicus Global Land Service provide an optical and thermal characterization of more than 4000 (optical) and over 1000 (thermal) inland water bodies . The production is based on Sentinel-3 data (OLCI and SLSTR) and Sentinel-2 (MSI) data for the currently ongoing NRT service. The products contain four (sets) of parameters: lake water surface temperature, lake water reflectance (all wavebands that are available after atmospheric correction), turbidity and a trophic state index in 11 classes derived from chlorophyll-a concentration. Production and delivery of the parameters are over set 10-day intervals. While turbidity, trophic state index and temperature are provided as 10-day averages, the lake water reflectance product contain the best representative spectrum of the covered time span in order to preserve the spectral reflectance vector. The products are available at 300m and 100m for the water quality parameters based on Sentinel-3 OLCI and Sentinel-2 MSI, respectively. Lake Surface Water Temperature (LSWT) is provided at 1km resolution. In all cases, a dedicated production process is used starting from Level 1 source products. The algorithms used to derive the optical lake water products are implemented in the Calimnos processing chain which contains procedures to pre-classify water into 13 optical water types associated with dedicated, tuned algorithms for chlorophyll-a and turbidity. Both LSWT and water quality algorithms represent the state-of-the-art for operational processing, including developments in the ESA SST_cci, Lakes_cci and GloboLakes projects.
Besides the NRT services based on OLCI and SLSTR, the full archives from MERIS and (A)ATSR are available for the time span 2002 – 2012. All products are publicly available via the Copernicus Global Land Service Website and Viewing Services to be easily accessible at the individual user or lake level, or by countries for further analysis and reporting. It is possible to provide tailored, aggregated products such as statistical records of chlorophyll-a concentration, algal bloom occurrence, percentage of surface water effected by eutrophic classes, etc.
The evolution of the services foresees additional parameters (such as chlorophyll-a concentration and floating cyanobacteria information) as well as extended coverage by the temperature products.
A key aspect of a Copernicus Service is its long term, sustained operational character. However, this is only assured if the delivered products meet user requirements and find wide uptake. We will present the current status of the lake water products, their interoperability and latest validation results. We will demonstrate how the products are applied to water management issues and how the products can be used in global scale analyses. As a prominent example, we will show how UNEP currently uses the products to inform per-country statistics on indicators under the Sustainable Development Goal 6.3.2 to harmonize information for SDG assessment and reporting.

Lakes are sentinels and integrators of environmental and climatic changes occurring within their watershed. The influence of climate change on lakes is becoming increasingly concerning worldwide. Understanding the complex behavior of lakes in a changing environment is essential to effective water resource management and mitigation of climate change effects. The increase in summer temperatures has been estimated at 0.34 °C per decade with lake specific parameters like morphology contributing to the diversity of response at the regional level. The frequency of heatwave events in Europe is increasing and the recent IPPC report on climate change estimated an over 90% likelihood that there will continue to be an increase in the frequency of heat extremes over the 21st century in Europe, especially in southern regions. In July 2019, a heatwave occurred in Europe with record daily maximum temperatures over 40 °C observed in several places. Temperatures were locally 6 to 8 °C higher than the average warmest day of the year for the period 1981–2010. Heatwaves can have implications for the water quality and ecological functioning of aquatic systems, and for example in Europe several of these events have been associated with increased phytoplankton blooms. Of particular concern is the predicted increase in potentially harmful summer blooms of cyanobacteria with combined pressures of climate change and eutrophication.
The ESA Climate Change Initiative (CCI) Lakes ECV Project (https://climate.esa.int/en/projects/lakes/) combines multi-disciplinary expertise to exploit satellite Earth Observation data to create the largest and longest possible consistent, global record of five lake climate variables: lake water level, extent, temperature, surface-leaving reflectance (e.g., chlorophyll-a and suspended solid concentrations), and ice cover. The first version of the database covers 250 globally distributed lakes with temporal coverage, depending on parameter, ranging from 1992 up to 2019. This is expanded to 2000 lakes in version 2. The ESA Lakes_cci dataset was found to be a key resource for examining the implications of heatwave events on lakes. We examined heatwave events for European lakes, focusing on the 2019 event. The response of lake chlorophyll-a concentration, a proxy of phytoplankton abundance, was dependent on the lake type, especially lake depth, stratification and trophic state. However, the timing of the heatwave event itself was also important in the type of response observed. In many cases, the effects of resulting storms that ended the heatwave were more discernable than the heatwave itself. For example, in some shallow lakes, following the storm, chlorophyll-a concentrations increased markedly and remained high for the duration of the summer. Comparing the high frequency WISPstation data (2018-2020) with the CCI dataset allows for detailed cross validation. Some of the rapid fluctuations visible from the satellite record are supported by the in situ data. In addition, utilizing the phycocyanin pigment estimates from the WISPstation and microscopic counts, showed how cyanophytes played a key role in the sudden increases and declines in chlorophyll-a in mid to late summer. Heatwaves and subsequent storms appeared to play an important role in structuring the phenology of the primary producers, with wider implications for lake functioning.

Abstract: This study is undertaken as part of the Lakes_cci project (ESA Climate Change Initiative), which aims to provide a multi-decadal, multi-sensor and global (over 2000 lakes) climate data record of water-leaving reflectance (Rw) and optical-biogeochemical water quality products. This presentation describes how the first non-interrupted multi-decadal satellite observations of global inland water quality have been created using the Calimnos processing chain, including the first per-pixel uncertainty characterization in a dynamic algorithm processing selection scheme for inland water bodies.
In the first phase of the Lakes_cci, which includes an initial dataset released for 250 lakes and a more recent improvement of spatiotemporal coverage to > 2000 inland water bodies, the primary aims for the Lake Water-Leaving Reflectance (LWLR) thematic Essential Climate Variable were to (1) improve the characterization of algorithmic uncertainties and provide uncertainty estimates with each pixel, and (2) to fill the observation gap between the MERIS and OLCI medium resolution sensors using MODIS-Aqua, where feasible.
We developed a method to produce estimates of Chlorophyll-a (Chla) satellite product uncertainty on a pixel-by-pixel basis within an Optical Water Type (OWT) classification scheme. This scheme helps to dynamically select the most appropriate algorithms for each satellite pixel, whereas the associated uncertainty informs downstream use of the data (e.g., for trend detection or modelling) as well as the future direction of algorithm research. Observations of Chla were related to 13 previously established OWT classes based on their corresponding normalized water-leaving reflectance (Rw), each class corresponding to specific bio-optical characteristics. Uncertainty models corresponding to specific algorithm - OWT combinations for Chla were then expressed as a function of OWT class membership score. Embedding these uncertainty models into a fuzzy OWT classification approach for satellite imagery allows Chla and associated product uncertainty to be estimated without a priori knowledge of the biogeochemical characteristics of a water body. Following blending of Chla algorithm results according to per-pixel fuzzy OWT membership, Chla retrieval shows a generally robust response over a wide range of class memberships, indicating a wide application range (ranging from 0.01 to 362.5 mg/m3). Low OWT membership scores and high product uncertainty identify conditions where optical water types need further exploration, and where biogeochemical satellite retrieval algorithms require further improvement. This work was conducted using MERIS observations, due to its long operation (2002-2012) and coincidence with matching in situ data in the LIMNADES (Lake Bio-optical Measurements and Matchup Data for Remote Sensing) database, and had been transferred to OLCI (2016-now) on Sentinel-3 considering its similarities in radiometric performance and waveband configuration with MEIRS.
To fill the gap between the MERIS and OLCI observation periods, independent validation and tuning of bio-optical algorithms was performed to facilitate the application of this procedure to MODIS. Product continuities between MERIS, OLCI and MODIS were then evaluated, and first attempts were made to remove the inter-sensor bias.
This work accompanies version 2.0 of the ESA Lakes_cci Climate Research Data Package, released in spring 2022. Version 1.1 of the dataset, covering 250 lakes, was published in July 2021 (https://catalogue.ceda.ac.uk/uuid/ef1627f523764eae8bbb6b81bf1f7a0a). Using the dataset published and combining several use cases, our presentation will also briefly showcase the global application of our products on phytoplankton phenology, climate change and decision making of lakes.

Lakes play a critical role in global climate regulation, carbon cycling, fresh water supply, fishing, and tourism, yet they remain vulnerable to climate change and anthropogenic disturbance. It is now widely acknowledged that climate warming influences lake functioning and ecosystem processes. However, significant gaps exist in our understanding of the potentially confounding interactions between climate forcing, water temperature and other water quality parameters. This study investigates the relationships between water temperature with chlorophyll-a (Chl-a) and turbidity in 250 lakes by analyzing long-term (2002-2019) satellite-retrieved data. We also elucidate the factors affecting these relationships. Daily lake median Lake Surface Water Temperature (LSWT) from ATSR, AASTR and AVHRR as well as derived Lake Water-leaving Reflectance (LWLR) products from MERIS (2002-2012) and Sentinel 3 OLCI (2016-2019) were extracted from the European Space Agency (ESA) Climate Change Initiative (CCI) Lake project (https://climate.esa.int/en/projects/lakes/). Time series of LSWT, Chl-a and turbidity were firstly detrended using the generalized additive model (GAM) fitting. A cross-correlation analysis was then carried out to investigate their patterns of relationship. We defined a total of six relationship patterns between LSWT and Chl-a/turbidity. The globally dominant pattern between LSWT and Chl-a was one where LSWT showed a negative and lagged relationship with Chl-a. In contrast, the globally dominant pattern between LSWT and turbidity was one where LSWT exhibited a positive lagged relationship following changes in turbidity. In total, 58% of the lakes showed negative relationships between LSWT and Chl-a, while 64% of the lakes showed positive relationships between LSWT and turbidity. The analysis of the factors influencing the observed relationships included lake area, depth, elevation, latitude, precipitation, wind speed, LSWT, Chl-a and turbidity. A boosted regression tree was used to analyze the relative influence of each factor on these relationships. It was found that LSWT and turbidity are the main factors influencing the relationship between LSWT and Chl-a, with relative influences of 16% and 15% respectively. Higher LSWT and lower turbidity conditions tend to lead to positive relationships between LSWT and Chl-a. For the relationship between LSWT and turbidity, lake area and Chl-a concentration are the main factors, with relative influences of 18% and 14% respectively. Lakes with smaller surface areas and with lower Chl-a concentrations tend to lead to positive relationships between LSWT and turbidity. This study will discuss sensitivities of the relationship between lake temperature and Chl-a/turbidity to climate change.

This study was supported by the European Space Agency Lakes Climate Change Initiative (ESA Lakes CCI+) project.

Remote sensing of the water-leaving radiance from airborne or spaceborne platforms requires the compensation for the absorption and scattering effects of the intervening atmosphere between the sensor and the surface. This process is known as “atmospheric correction” or “atmospheric compensation” (AC). Typical methods for AC are based on the assumption that the surface reflectance is roughly spatially homogeneous at a large spatial scale (> 30 km). This assumption is appropriate over large bodies of water, in regions away from the shore and when the spatial variations in the signal have no large contrast per waveband. However, this assumption is no longer valid in proximity to floating mats of opaque objects (macrophytes, constructions) and land, as those surfaces present high contrast with the water-leaving signal, typically in the near-infrared due to the high molecular absorption by pure water. These spatial heterogeneity with strong contrast will contribute to the observed signal over water due to the diffuse transmission of the atmosphere. This process, know as adjacency effect, biases the estimation of surface reflection under the assumption of spatial homogeneity.

In this study, we draw upon the theoretical and practical applications from the fields of atmospheric correction and atmospheric visibility to implement a sensor-agnostic adjacency correction in the frequency domain to the atmospheric correction code ACOLITE. The frequency domain approach greatly speeds up the computation for any given solution, allowing to use an iterative scheme to find the appropriate aerosol model and optical thickness. The processing chain can be summarized as:

1. For each aerosol model, iterate until the difference between the dark spectrum and the path reflectance is minimal, while keeping the fraction of negative VNIR pixels at a chosen threshold (0.1% of the scene):

1.1. Initial estimate of the aerosol optical thickness, under the assumption of spatially homogeneous surface reflectance;
1.2. Calculate the atmospheric point spread function;
1.3. Deconvolve at-sensor reflectance imagery and the point spread function;
1.4. Reconstruct the TOA reflectance without the upward diffuse contribution and estimate new dark spectrum;
1.5. Fit the corrected dark spectrum to the modeled path reflectance;

2. Select the aerosol model and optical thickness giving the lowest difference between the dark spectrum and the path reflectance.

The atmospheric Point Spread Functions (PSFs), per sensor waveband, for each atmospheric scatterer (maritime, continental, urban aerosols and Rayleigh) were calculated with a backward Monte Carlo code. The aerosol models and vertical profile from the 6SV code were used to calculate the atmospheric lookup tables in ACOLITE and the PSFs. Calculations were performed for pure scatterers and at a fixed optical thickness (aerosols) of 1 (unitless) and at variable surface pressures (Rayleigh). Results were then fitted to a model allowing to approximate the atmospheric PSF for variable mixing of atmospheric scatterers and surface pressure. These approximate equations are used in step 1.2 of the pseudocode above.

We evaluated the method applied to the Multispectral Instrument (MSI) onboard Sentinel-2A and B and to the Operational Land Imager (OLI) onboard Landsat 8, using field measurements from several small Belgian lakes made between 2017 and 2019. The results show a large increase in performance for the atmospheric correction in all bands and sensors, though the bias in the NIR wavebands remains large for low turbidity systems. The method showed better performance for OLI than for MSI, and this might be a consequence of the different spatial resolution and signal to noise ratio of the NIR bands of each sensor.

The root mean squared error (RMSE) when using the adjacency correction was on average 4 times (OLI/Landsat 8) and 2 times (MSI/Sentinel-2) lower than when the AC assumes spatial homogeneity.

Creating multi-mission satellite-derived water quality (WQ) products in inland and nearshore coastal waters is a long-standing challenge due to the inherent differences in sensor spectral and spatial sampling as well as in their radiometric performance. This research extends a recently developed machine-learning (ML) model, i.e., Mixture Density Networks (MDNs) to the inverse problem of simultaneously retrieving WQ indicators, including chlorophyll-a (Chla), Total Suspended Solids (TSS), and the absorption by Colored Dissolved Organic Matter at 440 nm (a_cdom (440)), across a wide array of aquatic ecosystems. We use an in situ database to train and optimize MDN models developed for the relevant spectral measurements (400 – 800 nm) of the Operational Land Imager (OLI), MultiSpectral Instrument (MSI), and Ocean and Land Colour Instrument (OLCI) aboard the Landsat-8, Sentinel-2, and Sentinel-3 missions, respectively. Our performance assessments suggest varying degrees of improvements with respect to second-best algorithms, depending on the sensor and WQ indicator (e.g., 68%, 75%, 117% improvements for Chla, TSS, and a_cdom (440), respectively from MSI-like spectra). Map products are demonstrated for multiple OLI, MSI, and OLCI acquisitions to evaluate multi-mission product consistency across broad spatial scales. Overall, estimated TSS and a_cdom (440) from these three missions are consistent within the uncertainty of the model, but Chla maps from MSI and OLCI are more accurate than those from OLI. Through the application of two different atmospheric correction processors to OLI and MSI images, we also conduct matchup analyses to quantify the sensitivity of the MDN model and best-practice algorithms to uncertainties in remote sensing reflectance products. The analysis indicates our model is less or equally sensitive to these uncertainties compared to other algorithms. Recognizing their uncertainties, MDN models can be applied as a global algorithm to enable harmonized retrievals of Chla, TSS, and a_cdom (440) in various aquatic ecosystems.

As part of the Copernicus long-term scenario, ESA is planning the development of the Sentinel-1 Next Generation (NG) mission. It’s main goal is to ensure the C-band data continuity beyond the next decade (2030) in support of the operational Copernicus services that are routinely using Sentinel-1 data. In addition, the enhanced capabilities of Sentinel-1 NG compared to Sentinel-1, will support novel imaging capabilities and enable further development and improvement of operational applications.
The Sentinel-1 NG mission addresses three cross-cutting categories of requirements. These are:
1. Continuity which are requirements linked to the continuation of current observations in order to establish long and reliable time series of measurements. The compatibility of different sources and data streams has to be ensured across time and space.
2. Enhanced continuity which are requirements for improvements in current observations in terms of accuracy, quality in general, and temporal and spatial resolution.
3. Observation gaps which are requirements corresponding to parameters not yet observed by Copernicus. These are not necessarily the observations to be provided with the highest priority. Some of them cannot be provided by instruments embarked on satellites given the current state of knowledge, others could be secured through international collaboration and data exchange rather than committing limited resources to satellite mission developments.
The primary goal of Sentinel-1 NG is to ensure the continuity of Copernicus and other important services that rely on Sentinel-1 as a primary source of information to deliver the service. Continuity in this context is defined as the continuity of services and applications. In addition to ensuring continuity of services, the S1 NG mission and satellites shall also address the enhanced continuity of current services as well as possible new areas of development or evolution of services responding to the evolving capabilities of the user community.
Sentinel-1 NG will directly address the information needs of numerous services including inter alia:
- The nascent European Ground Motion Service (EGMS) which is now part of the Copernicus Land Monitoring Service’s product portfolio. EGMS is a service that aims to provide consistent, regular, standardized, harmonized and reliable information regarding natural and anthropogenic ground motion phenomena over Europe and across national borders, with millimeter accuracy.
- The Copernicus Emergency Management Service (EMS) providing enhanced continuity and improved geospatial information to assess the impact of natural and man-made disasters all over the world.
- The Copernicus Land Monitoring Service (CLMS) where Sentinel-1 NG is expected to support enhanced continuity and service evolution are the many applications centred on land cover mapping and soil moisture information.
- Maritime Safety and Security including continuity with respect to existing Maritime Surveillance Services for oil spill detection and new capabilities to support vessel monitoring leveraging the higher resolution and wide swath of Sentinel-1 NG.
- Monitoring of the Arctic seas through enhanced continuity in monitoring ice and the change of ice over time, supporting transport and human activities with key environmental information for safety and environmental protections purposes and security applications.
As can be seen above Sentinel-1 NG represents a multi-purpose SAR mission whose design is required to support multiple applications and services. The driving mission requirements are the combination of higher resolution, wider swath and more frequent coverage with respect to the current generation of Sentinels. As an example a spatial resolution of 5m x 5m is being targeted compared to the 20m x 5m typical of the interferometric wide swath mode of Sentinel-1.
The technical concept of Sentinel-1 NG is currently being studied by European industry within the context of a Phase-AB1 study. Special consideration is given to the compatibility of the mission with the current Sentinel-1 and future ROSE-L constellation with harmonised coverage, e.g. on another node of the same orbital plane or observing the same ground area few seconds apart (typically 60 seconds or less).

The Copernicus Sentinel-2 Next Generation (NG) mission will be the new European reference multi-spectral imaging mission supporting Copernicus Services related to land, coastal areas, climate change, emergency management, and security. In continuity with Sentinel-2, Sentinel-2 NG will be a workhorse for the Earth Observation (EO) and geospatial scientific communities, as well as a crucial data provider for the Copernicus EO downstream market.

Scheduled for launch in the early 2030s, the Sentinel-2 NG space segment and its mission specific ground segment will ensure enhanced continuity of service to the current Sentinel-2 data in order to meet new and emerging user needs. This will be achieved thanks to the enhanced observational capabilities of the Advanced Multi-Spectral Imager (AMSI), allowing a higher spatial resolution while concurrently maintaining (as a minimum) the same image quality and accommodating a larger number of spectral bands compared to Sentinel-2. A lower revisit time compared to the current generation of Sentinel-2 is also being targeted in response to the users’ request. During pre-Phase 0 and Phase 0, ESA – together with the Sentinel-2 NG Ad Hoc Expert Group (AHEG) – defined a baseline set of spectral bands for potential implementation in AMSI. These spectral bands reflect specific requirements expressed by the user community, the relevant Copernicus services and the Sentinel-2 NG AHEG. Next to the current Sentinel-2 MSI heritage bands, some new and additional bands that can enable new applications or enhance existing ones are being considered. Complementarity and coordination with other Copernicus missions expected to fly in the same time frame as Sentinel-2 NG (e.g. CHIME, LSTM, Sentinel-3 NG) is being strongly pursued. Moreover, synergies with the future NASA/USGS Landsat Next mission are being considered.

In addition to the multispectral bands, the user requirement analysis showed that some applications may benefit from the availability of a very high resolution (VHR) or panchromatic (PAN) band. Although the addition of a PAN band onboard the Sentinel-2 NG satellite has been discarded from the Phase 0 study, the inclusion of VHR data, i.e. with a spatial sampling distance below 2.5 m, within the Sentinel-2 NG overall mission architecture is being considered as a relevant feature. Specific user and mission requirements to pave the way for a combined use of AMSI measurements and VHR data are currently being identified.

The Sentinel-3 Mission provides an essential satellite altimetry data set for the Copernicus services over the global ocean, coastal zones, sea and land ice and inland waters in support of a large number of end user applications. Four Sentinel-3 satellites (two currently on orbit and two replacements to be launched in the coming years) operating in a sun-synchronous near polar orbit provide an unprecedented and unbroken time series of satellite altimetry measurements from 2016-2035.

Considering the User needs expressed by the European Commission and inputs from an independent Mission Advisory Group are documented in the S3-NG-T Mission Requirements Document. The aim of the Copernicus Next Generation Sentinel-3 Topography (S3NG-T) Mission is to guarantee baseline continuity of existing Copernicus Sentinel-3 nadir-altimeter measurements in the 2030-2050 time-frame while enhancing measurement performance.

The primary objectives of the S3NG-T mission are to:

PRI-OBJ-1. Guarantee continuity of Sentinel-3 topography measurements for the 2030-2050 time frame with performance at least equivalent to Sentinel-3 in-flight performance (‘baseline mission’).
PRI-OBJ-2. Respond to evolving user requirements and improve sampling, coverage and revisit of the Copernicus Next Generation Topography Constellation (S3NG-T and Sentinel-6NG) to "≤ 50" km and "≤ 5" days in support of Copernicus User Needs.
PRI-OBJ-3. Enhance sampling coverage, revisit and performance for Hydrology Water Surface Elevation measurements in support of Copernicus Services.
PRI-OBJ-4. Respond to evolving user requirements and enhance topography Level-2 product measurement performance.

The secondary objectives of the S3NG-T mission are to:

SEC-OBJ-1. Provide directional wave spectrum products that address evolving Copernicus user needs.
SEC-OBJ-2. Provide new products (e.g. sea surface height gradients and river reach averaged gradients, river width and water area. that address evolving Copernicus user needs.

A coordinated constellation of spacecraft is required to provide enhanced continuity to the Sentinel-3 Mission to meet sampling at "≤ 5" days, "≤ 50" km (25 km wavelength) between 81.5 degrees north and south of the equator - regardless of the satellite technologies employed. Such a constellation requires a reference that allows satellite to be used in synergy with each other to satisfy User needs without bias discrepancies. For the S3NG-T mission, it is assumed that a reference satellite mission (e.g. Copernicus Sentinel-6/NG) will be available in orbit that is designed for this purpose providing a common reference measurement for all S3NG-T satellites and with excellent knowledge of measurement stability. Additional third party altimeter missions may also provide additional data although their launch or access to their data cannot be guaranteed.

Continuity of Sentinel-3 measurements can be guaranteed using a number of different technical solutions with potential enhancements in terms of coverage, sampling, calibration stability, system complexity and size amongst others. The best implementation approach for the S3NG-T mission shall be based on "fitness for purpose to provide enhanced continuity to Sentinel-3 topography measurements" determined by compliance to mission requirements, maturity of technical heritage, technical feasibility/readiness, scientific readiness and maturity, development schedule, risk, cost and programmatic arrangements. Based on the S3NG-T Phase-0 activities and other studies conducted by ESA over the last 5 years the most likely scenarios to implement S3NG-T include the following:

• Scenario-1: Replacement of Sentinel-3C and Sentinel-3D using a constellation of 2-n nadir-pointing altimeters.
• Scenario-2: Implementation of 2..n swath altimeter including a nadir altimeter.
• Scenario-3: A hybrid approach including both nadir pointing altimeter and swath altimeter satellites

This paper will review the current status of the S3NG-T Mission following the completion of Phase-0 studies and on-going Phase A/B1 studies at ESA.

Sustained observations in the visible and infrared domain are one of the pillars of the Copernicus Programme. The vast amount of data collected by the current Sentinel-3 optical payloads, i.e. the Ocean and Land Colour Imager (OLCI) and the Sea and Land Surface Temperature Radiometer (SLSTR), provide crucial input to a number of Services including the Marine Service (CMEMS) and Land Monitoring Service (CLMS)
To ensure continuity of service for Copernicus, the next generation (NG) of the Copernicus Space Component (CSC) of the Sentinel-3 mission is foreseen to be launched in the 2032 time horizon. The Next Generation of Sentinel-3 Optical (Sentinel-3 NG Optical) will address the evolution of the optical payloads OLCI and SLSTR , while the radar payload will be addressed by a different mission (Sentinel-3 and Sentinel6 NG Topography).
The Sentinel-3 NG Optical will aim at the delivery of crucial observations both over oceans and over land . In the oceanic domain these observations are needed to constrain and drive global and local ocean assimilation models and coupled ocean/atmosphere assimilation models. For this, Sentinel-3 Optical NG will deliver as primary mission objectives:
• Continuity of sea colour data, at least at the level of the quality of the current generation of Sentinel-3 OLCI
• Continuity of sea surface temperature, at least at the level of the quality of the current generation of Sentinel-3 SLSTR
Observations of moderate resolution and wide swath from this mission will also provide information needed to derive global land products and feed the related services with data. These are:
• Continuity of land surface colour at least at the level of quality of the current generation of Sentinel-3 OLCI
• Continuity of land surface temperature at least at the level of quality of the current generation of Sentinel-3 SLSTR
• Continuity of vegetation products, based on synergetic measurements from optical instruments at least at the level of quality of the current generation of Sentinel-3

Ensuring continuity goes hand in hand with the technical evolution of the OLCI and SLSTR instruments, in order to improve the products; this in turn leads to enhanced services or new services, and also enables R&D into new applications. In this presentation we will review the key scientific and service requirements of the mission, illustrate the advanced technical concepts that are currently being evaluated as part of Phase 0 of the mission development, and discuss some of the expected applications.

As part of the Copernicus component of the European Union (EU) Space Programme, the European Space Agency (ESA) and EUMETSAT are preparing for the expansion of the Copernicus Space Infrastructure with new observation capabilities for monitoring greenhouse gas emissions, marine and polar areas. This includes in particular the development of an Anthropogenic CO2 Monitoring mission (CO2M), a Polar Ice and Snow Topography Mission (CRISTAL), and a Polar Ice and Ocean Imaging Microwave Radiometer Mission (CIMR).
Through a Contribution Agreement with the European Union, EUMETSAT was entrusted to contribute to the development of those three missions, taking up a different role for each of them.
EUMETSAT contributes to the development of a significant part of the CO2M ground segment and is responsible for the routine operations of the Anthropogenic CO2 Monitoring mission (CO2M). EUMETSAT will undertake the day-to-day routine satellite operations of CO2M and the continuous processing, monitoring, validation and, where needed, vicarious calibration of the payload data-products and their operational dissemination to users. ESA is contributing with the development of the space-segment and the remaining CO2M ground segment elements. ESA will also perform the satellite in-orbit verification activities and operate the satellites in this phase.
In the case of CRISTAL, when the mission is confirmed by the European Commission, EUMETSAT will be responsible for the deployment of data processing chains for global ocean products, including their validation aspects, in synergy with Sentinel-3/-6. This includes L1 and L2, as well as L2P/L3 products over global ocean.
In the case of CIMR, also subject to confirmation from the Commission, EUMETSAT will be responsible for the deployment of data processing chains for generating L2 products over the global ocean, in order to support marine and weather applications, extracting the associated geophysical parameters, in synergy with other relevant missions operated by EUMETSAT
This presentation provides an overview of the main logical elements of the CO2M operational ground segment. In particular, we provide an overview of the key parameters and products, which can be expected from CO2M and point to specific challenges for a future operational CO2 monitoring system.
In addition, the presentation addresses the activities within EUMETSAT in preparation for the potential CRISTAL and CIMR missions.

BRIX-2 stands for Second Biomass Retrieval Inter-comparison eXercise and represents a joint effort between ESA and NASA to intercompare algorithms specifically for biomass mapping using current and future spaceborne missions. The exercise aims at using Synthetic Aperture Radar (SAR) at P-Band and L-band, and LIDAR datasets acquired as part of the ESA and NASA AfriSAR joint-campaign in support of the upcoming ESA’s BIOMASS [1] mission, of the upcoming NASA-ISRO SAR (NISAR) mission [2], and of the current NASA Global Ecosystem Dynamics Investigation (GEDI) mission [3].

The objectives of BRIX-2 are:
1. Provide an objective, standardized comparison and assessment of biomass retrieval algorithms developed for the BIOMASS, NISAR and GEDI missions, and fusion of these mission datasets.
2. Establish a forum to involve scientists in the development of retrievals that have so far not been part of the Biomass community.
3. The adoption of vetted validation standards and methods to compare biomass estimates to reference datasets (e.g. field plots or airborne lidar biomass maps).
4. Collect inputs from the biomass user and scientific community on data formats and characteristics towards the generation of Analysis Ready Data.

These objectives shall be achieved by making available standardized test cases (based on airborne campaign and spaceborne simulated data), inviting the scientific community to develop and apply retrieval algorithms based on this test case, and finally compare and evaluate the performance of submitted results [4].

For the purpose of an objective algorithm evaluation, the exercise was based on ESA-NASA Multi-Mission Algorithm and Analysis Platform (MAAP) [5]. This analysis platform is a virtual, open and collaborative environment for the processing, analysis and sharing of data and development and sharing of algorithms. The MAAP provides a common platform with computing capabilities co-located with data as well as a set of tools and algorithms developed to support this specific field of research.

Participants were invited to upload their code with a mandatory permissive open-source license to the MAAP (or develop it on the MAAP) and run it on the MAAP using the predefined campaign datasets.

The first results of the this inter-comparison exercise will be presented.


REFERENCES
[1] T. Le Toan, S. Quegan, M. Davidson, H. Balzter, P. Paillou, K. Papathanassiou, S. Plummer, F. Rocca, S. Saatchi, H. Shugart and L. Ulander, “The BIOMASS Mission: Mapping global forest biomass to better understand the terrestrial carbon cycle”, Remote Sensing of Environment, Vol. 115, No. 11, pp. 2850-2860, June 2011.

[2] P.A. Rosen, S. Hensley, S. Shaffer, L. Veilleux, M. Chakraborty, T. Misra, R. Bhan, V. Raju Sagi and R. Satish, "The NASA-ISRO SAR mission - An international space partnership for science and societal benefit", IEEE Radar Conference (RadarCon), pp. 1610-1613, 10-15 May 2015.

[3] https://science.nasa.gov/missions/gedi

[4] « Biomass Retrieval Inter-comparison eXercise #2: BRIX-2 Protocol », version 2.1, 6 August 2021, https://liferay.val.esa-maap.org/documents/portlet_file_entry/35530/BRIX-2+Protocol+V2.1.pdf/1d887f7c-5a15-9725-0ec8-3a50292f0010?download=true]

[5] Albinet, C., Whitehurst, A.S., Jewell, L.A. et al. A Joint ESA-NASA Multi-mission Algorithm and Analysis Platform (MAAP) for Biomass, NISAR, and GEDI. Surv Geophys 40, 1017–1027 (2019). https://doi.org/10.1007/s10712-019-09541-z

Accurate mapping of forest aboveground biomass (AGB) is critical for carbon budget accounting, sustainable forest management as well as for understanding the role of forest ecosystem in the climate change mitigation. In this study, spaceborne Global Ecosystem Dynamics Investigation (GEDI) LiDAR and Sentinel-2 multispectral data were used in combination with elevation and climate data to produce a wall-to-wall AGB map of Australia that is more accurate and with higher spatial and temporal resolution than what is possible with any one data source alone. Specifically, the AGB density map was produced that covers the whole extent of Australia at 200m spatial resolution for the Austral winter (June-August) of 2020. To produce this map Copernicus Sentinel-2 composite, GLO-90 Digital Elevation Model (DEM) and long-term climate variables were trained with samples from the GEDI Level 4A product.
From GEDI Level 4A data available within Australia between June – August 2020, all measurements not meeting the requirements of L4A product quality, and those with degraded state of pointing or positioning information and an estimated relative standard error in GEDI-derived AGB exceeding 50% were rejected. Seasonal Sentinel-2 composite was generated using a Sentinel-2 Global Mosaicking (S2GM) algorithm and was further used to calculate Normalized Difference Spectral Indices (NDSIs) from all spectral bands. Similarly, DEM was used to calculate aspect, roughness, slope, Topographic Position Index and Terrain Ruggedness Index. Finally, climate variables consisted of average precipitation, radiation as well as minimum and maximum temperatures calculated between 1970-2020.
The boosting tree machine learning model was applied to predict wall-to-wall AGB density map. For each 200m × 200m cell the number of available GEDI measurements was calculated and models were built based on average AGB density of cells containing > 5 GEDI measurements. Up to ≈62,000 cells, each 200m × 200m, were used to train predictive machine learning models of AGB density. The predictive performance of models based on both satellite imagery only (single-data source) and a fusion of satellite imagery with elevation and climate data (multi-data source) was compared. Bayesian hyperparameter optimization was used to identify the most accurate Light Gradient Boosting Machine (LightGBM) model using 5-fold cross-validation.
The multi-data source approach had a substantially higher accuracy (coefficient of determination (R2) increase of up to 0.1, root-mean-square error (RMSE) decrease of up to 7 Mg/ha and root-mean-square percentage error (RMSPE) decrease of up to 17%) as compared to the single-data source approach. The single-data source analysis based on only Sentinel-2 imagery resulted in AGB density predicted with the R2 of 0.68-0.75, RMSE of 40-46 Mg/ha and RMSPE of 47-69%. Model performance improved with the addition of DEM and climate information: AGB density prediction with R2 of 0.77-0.81, RMSE of 35-40 Mg/ha and RMSPE of 41-52%. Using a SHapley Additive exPlanations (SHAP) approach to explain the output of LightGBM models it was found that Sentinel-2 derived NDSIs using Red Edge and Short-wave Infrared bands were the most important in predicting seasonal AGB density.
Similar model performance is expected for annual prediction of AGB density at a finer resolution (e.g. 100m) due to higher density of GEDI measurements. This research highlights methodological opportunities for combining GEDI measurements with satellite imagery and other environmental data toward seasonal AGB mapping at the regional scale through data fusion.

TomoSense is an ESA-funded campaign over a temperate forest in support of future SAR mission concepts at P-, L- and C-band. This paper gives the first results from analysing the relationship between P-band tomographic SAR (TomoSAR) data and above-ground biomass (AGB) including ground slope effects. The results are important for the upcoming BIOMASS mission, in particular for temperate forest AGB estimation.

Airborne P-band SAR data over the Kermeter region in Germany was acquired and processed by MetaSensing. TomoSAR processing was performed by Politecnico di Milano by combining 56 SAR flight tracks (28 in each direction – NW and SE). The flights were performed in the mornings on the 22 and 23 July 2020. The SE track acquisitions were affected by RF interference, and only the NW tracks were used for this analysis. An AGB map based on airborne laser scanning (ALS) and 80 in-situ plots, each of a size of 500 m², was provided by CzechGlobe. In addition, a digital elevation model (DEM) over the area provided input for the assessment of topographical effects.

A mix of all forest biotypes (dominant species were beech and spruce, otherwise oak, pine and birch) showed an AGB RMSE of 15 % for VV polarization, 17 % for HV and 29 % for HH relative the AGB map. The results obtained were based on 397 areas of 0.5 ha size, limiting the surface slope angle to be less than 20°, and integrating the TomoSAR intensity from 20 m to 30 m in height. TomoSAR vertical profiles in this height interval were found to have the highest sensitivity to AGB. The corresponding AGB RMSE for integrating the intensity of the vertical profiles over the full height was 30 % in VV, 57 % in HV, whereas HH did not show a clear sensitivity to AGB.

Influence from topography on the AGB sensitivity was observed for increasing surface slope angles. Limiting the surface slope angle to below 20° improved the AGB RMSE from 22 % to 15 % in VV and from 24 % to 17 % in HV, when observing a mix of all forest biotypes. In the ground-range direction, negative slopes (facing away from the radar) decrease the TomoSAR intensity integrated from 20 m to 30 m whereas positive slopes (facing toward the radar) increase it. This effect is likely due to a combination of varying canopy attenuation and incidence angle dependent trunk scattering. A degradation of the AGB sensitivity was also observed for increasing slopes in the azimuth direction due to increased intensity variance per AGB interval.

These results show that TomoSAR, which will be available from BIOMASS, is a promising technique for estimating AGB in temperate mixed forests. Similar observations over the region at L- and C-band are currently being analysed.

BIOMASS [1] is ESA's seventh Earth Explorer, scheduled for launch in 2023. It will collect unprecedented information about forests thanks to the first spaceborne P-band Synthetic Aperture Radar (SAR), and featuring full polarimetry. The 435 MHz carrier frequency and 6 MHz bandwidth allow maximum sensitivity to the woody elements of the trees, while complying with ITU regulations and avoiding excessively strong ionospheric effects. Its acquisition cycle is designed to acquire globally (subject to Space Object Tracking Radars restricitions, excluding North America and Europe) and in a multi-baseline repeat pass interferometric configuration. The repeat cycle is set to 3 days, guaranteeing coherence for interferometric and tomographic processing.
After the initial Commissioning Phase, dedicated to instrument and antenna calibration, the experimental Tomographic Phase will achieve one global coverage in 14/16 months. This will allow mapping forests in 3 dimensions by collecting stacks of 7 acquisitions over 18 days for each location, in ascending and descending configurations. During Tomographic Phase it will be also possible to derive a sub-canopy Digital Terrain Model (DTM) [3]. This is fundamental to reject terrain contribution in interferometric processing [5], as terrain acts as a nuisance and disturbs biomass retrieval. The remainder of the mission is the main Interferometric Phase of about 5 years duration, with global coverage achieved in 7/9 months. In this case stacks are collected in dual baseline configuration over 6 days. Coverage will be built up successively, with the successive tomographic or interferometric stacks adding coverage to adjacent areas. Given the complex orbital pattern achieving global coverage over several months, significant environmental changes will occur that the estimation techniques must handle.
In this contribution we describe the Level-2 processing algorithms to estimate Forest Disturbance (FD), Forest Height (FH) and Above Ground Biomass (AGB) products from BIOMASS data [2]. The Level-2 processor requires phase calibrated stacks generated by the BIOMASS interferometric processor and the DTM estimated in the Tomographic Phase [6]. FD, FH and AGB products are generated at each global coverage during the mission lifetime mapping not only forest characteristics but also changes. The processing implements state-of-the-art polarimetric-interferometric techniques allowing to reject terrain signal and focus on canopy scattering. This is supported by strong evidence that in tropical forests the backscatter from the canopy region 25-35m above the ground is highly correlated with the total AGB, which can be exploited using the full power of tomography or interferometry. The actual Level-2 processor software implementation is also briefly presented, along with preliminary results on airborne campaign data.
In particular, AGB estimation results will be shown on BIOMASS-like acquisitions emulated from tropical forests campaign data. The AGB estimation performance is observed to depend on the AGB range and degrades when ground topography is significant. Good performance is achieved when the AGB interval is large (> 400 t/ha) and the average is in the interval 200–250 t/ha [4]. The algorithm is observed to be capable of achieving a relative RMSD of 20% with respect to in situ data using only few calibration points where reference AGB is available, although retrieval accuracy was observed to depend significantly on the quality of the available calibration points. Efforts are now focused on designing the global AGB estimation scheme for BIOMASS, especially with regards to calibration and validation AGB to be used.

[1] T. Le Toan, S. Quegan, M.W.J. Davidson, H. Balzter, P. Phaillou, K. Papathanassiou, S. Plummer, F. Rocca, S. Saatchi, H. Shugart, L. Ulander, “The BIOMASS mission: Mapping global forest biomass to better understand the terrestrial carbon cycle,” Remote Sensing of Environment, vol. 115, pp. 2850-2860, Jun. 2011
[2] Banda, F.; Giudici, D.; Le Toan, T.; Mariotti d’Alessandro, M.; Papathanassiou, K.; Quegan, S.; Riembauer, G.; Scipal, K.; Soja, M.; Tebaldini, S.; Ulander, L.; Villard, L. The BIOMASS Level 2 Prototype Processor: Design and Experimental Results of Above-Ground Biomass Estimation. Remote Sens. 2020, 12, 985.
[3] Mariotti D’Alessandro, M.; Tebaldini, S. “Digital Terrain Model Retrieval in Tropical Forests Through P-Band SAR Tomography” IEEE Transactions on Geoscience and Remote Sensing, vol. 57, no. 9, pp. 6774-6781, Sept. 2019. doi: 10.1109/TGRS.2019.2908517
[4] Soja, M., Quegan, S., Mariotti d’Alessandro, M. Banda, F. Scipal, K., Tebaldini, S. Ulander, L.M.H. “Mapping above-ground biomass in tropical forests with ground-cancelled P-band SAR and limited reference data”, Remote Sens. Environ. Volume 253, February 2021, 112153
[5] M. Mariotti d’Alessandro, S. Tebaldini, S. Quegan, M. J. Soja, L. M. H. Ulander and K. Scipal, "Interferometric Ground Cancellation for Above Ground Biomass Estimation," in IEEE Transactions on Geoscience and Remote Sensing, vol. 58, no. 9, pp. 6410-6419, Sept. 2020
[6] M. Pinheiro at al., “BIOMASS DEM Product Prototype Processor”, EUSAR 2021

Trees and shrubs (hereafter collectively referred to as trees) both inside and outside forests, are the basis for the functioning of tree-dominated ecosystems, and are regularly monitored at country scale via forest inventories. However, traditional inventories and large-scale forest mapping projects are expensive, labour-intensive and time-consuming, resulting in a trade-off between the details recorded, spatial coverage, accuracy, regularity of updates, and reproducibility. Also, forest inventories typically do not account for individual trees outside forests, although these trees play a vital role in sustaining communities through food supply, agricultural support, among other benefits. Moreover, the alarming rate of tree cover loss resulting from different natural and human-induced processes has brought both political and economic motives to attract efforts for landscape restoration especially in Africa. Nevertheless, currently, there is no accurate and regularly updated monitoring platform to track the progress and biophysical impact of such ongoing initiatives. Recent approaches counting trees in satellite images in Africa used very costly commercial images, were limited to isolated trees in savannas excluding small trees, and did not cover other complex and heterogeneous ecosystems such as forests. Here, we make use of novel deep learning techniques and publicly available aerial photographs, and introduce an accurate and rapid method to map the crown size, number of trees inside and outside forests, and corresponding carbon stock, regardless of tree size and ecosystem types in Rwanda. The applied deep learning model follows a UNet architecture and was trained using 67,088 manually labeled tree crowns. We mapped over 200 million individual trees in forests, farmlands, wetlands, grasslands, and urban areas, and found about 67.2% of the mapped trees outside forests. An average tree density of 94.6 and 70.8 trees per ha, and average crown size of 38.7 m2 and 15.2 m2 were mapped inside and outside forests, respectively. In savannas we found 64 trees per ha with an average crown size of 15.6 m2. In farmlands we found 79.6 trees per ha with an average crown size of 16.3 m2. We expect methods and results of this kind to become standard in the near future, enabling tree inventory reports to be of unprecedented accuracy.

The use of airborne laser scanning (ALS) data to estimate and map structure-related forest inventory variables, such as forest aboveground biomass, has strongly increased in the last decades and has even become operational in some countries. The development of new machine learning methods, including deep learning-based approaches, and the fine-tuning and validation of already existing methods for deriving these variables from ALS data require extensive datasets of forest inventory measurements on a single-tree level and corresponding ALS data.
Virtual laser scanning (VLS) is a time- and cost-efficient alternative to acquiring such data in the field. We present a framework for virtual laser scanning that combines forest inventory data with a tree point cloud database and an open-source laser scanning simulation framework.
Synthetic 3D representations of forest scenes are created based on forest stand information that can be derived from a forest inventory on a single-tree level or a forest growth simulator. For each tree in the forest stand, a 3D tree model of matching species, height, and crown diameter is inserted into the synthetic forest scene. Single-tree point clouds extracted from real ALS data are used as tree models. Laser scanning of the synthetic forest scene is simulated using the Heidelberg LiDAR Operations Simulator HELIOS++.
We investigate the performance of the VLS simulations using ALS and forest inventory data collected from six 1-ha plots in temperate forests in Southwest Germany. VLS is performed with the same acquisition settings as in the real ALS campaign. For a comparison of different tree model types, the synthetic forest stands are created using closely matching real tree point clouds and using two types of simplified tree models with cylindrical stems and spheroidal crowns. The simulated ALS point clouds are compared with the real ALS point clouds both qualitatively, based on their visual appearance in cross-sections, and quantitatively, based on the height distribution of the returns and several point cloud metrics. To assess the potential of the synthetic data in an application, they are used as training data for forest biomass models, which are then applied on the real ALS data.
Our validation confirms that the presented workflow can be used to generate synthetic ALS datasets, which are sufficiently realistic for many typical applications. The VLS approach can reproduce the relative height distribution of returns of real ALS data. The comparison of different tree model types reveals that the visual appearance of the synthetic forest scenes is much more realistic for the real tree point clouds than for the simplified tree models. However, the differences between the real tree point clouds and the simplified tree models are less pronounced in the quantitative analysis. Our findings suggest that for the wall-to-wall mapping of forest aboveground biomass, synthetic forest scenes composed of simplified tree models can be as suitable as synthetic forest scenes composed of single tree point clouds to create training datasets.

Due to the rapidly changing climate in the polar regions, the thawing of permafrost soil leads to disturbances like retrogressive thaw slumps (RTS). As these disturbances can further destabilize permafrost and cause damage to existing infrastructure, they require continuous monitoring. They often appear in clusters and are comparably small in size, rarely exceeding 10 ha in size, thus calling for high resolution remote sensing data. Recently launched satellite platforms and publically available datasets (e.g. ArcticDEM) enable this high resolution monitoring from space. However, the vast extent of permafrost areas renders manual analysis infeasible and calls for an automated approach to mapping these disturbances. We present such an automated approach using deep learning.
As deep neural networks generally require large annotated training datasets, we first create a dataset of manually annotated thaw slumps on PlanetScope satellite imagery. This dataset contains imagery from 2018 and 2019 and spans six areas of interest in Canada (Banks Island, Herschel Island, Horton Delta, Tuktoyaktuk Peninsula) and Russia (Kolguev Island, Lena River). The overall spatial extent of the dataset is around 900km² and covers 2172 individual thaw slumps. In extensive pre-processing steps, the PlanetScope optical imagery was enriched by ArcticDEM-derived topography data like relative elevation and slope, as well as tasseled cap time series trends derived from Landsat imagery. Finally, the data is cut into smaller spatial patches to facilitate the training of deep neural networks.
Using this dataset as a basis, we explore deep learning approaches to automated RTS mapping. The task of pixel-wise classification is well researched in computer vision, therefore our approach builds on existing frameworks for semantic segmentation. However, these models were built on assumptions from a different field, with some expecting readily pre-trained backbone networks. Therefore, it is not straight-forward to predict which model will perform best at RTS mapping, calling for a careful evaluation of the models on our dataset. To this end, various configurations of the UNet, UNet++ and DeepLabv3 segmentation models are trained and evaluated. Given the spatial sparsity of the target features, combinations of two training protocols are examined. In sparse training, the model is trained only on patches that contain target features, while in full training, the model is trained on all patches. In our experiments, the best performing combination was a long phase of sparse training followed by a short phase of full training.
In order to evaluate the general performance and spatial transferability of the models, we perform a regional cross-validation on the six study sites, where 5 are used for training and the remaining site is used to evaluate the model performance. The observed model accuracy is promising in some of the validation regions (e.g. Lena, Horton, Kolguev), while others still show room for improvement (Banks Island, Tuktoyaktuk). This can be attributed due to the large variance in geomorphological features between the evaluated regions, where some are easier to generalize to, while others pose a harder challenge for the models.

The automatic detection and tracking of mesoscale ocean eddies, the ‘weather of the ocean’, is a well-known task in oceanography. These eddies have horizontal scales from 10 km up to 100 km and above. They transport water mass, heat, nutrition, and carbon and have been identified as hot spots of biological activity. Monitoring eddies is therefore of interest among others to marine biologists and fishery.

Recent advances in satellite-based observation for oceanography such as sea surface height (SSH) and sea surface temperature (SST) result in a large supply of different data products in which eddies are visible. In radar altimetry observations are acquired with repeat cycles between 10 and 35 days and cross-track spacing of a few 10 km to a few 100 km. Therefore, ocean eddies are clearly visible but typically covered by only one ground track. In addition, due to their motion, eddies are difficult to reconstruct, which makes creating detailed maps of the ocean with a high temporal resolution a challenge. In general, they are considered a perturbation, and their influence on altimetry data is difficult to determine, which is especially limiting for the determination of an accurate time-averaged dynamic topography of the ocean.

Due to their spatio-temporal dynamic behavior the identification and tracking are challenging. There is a number of methods that have been developed to identify and track eddies in gridded maps of sea surface height derived from multi-mission data sets. However, these procedures have shortcomings since the gridding process removes information that is valuable in achieving more accurate results.

Therefore, in the project EDDY carried out at the University of Bonn we intend to use ground track data from satellite altimetry and - as a long-term goal - additional remote sensing data such as SST, optical imagery, as well as statistical information from model outputs. The combination of the data will serve as a basis for a multi-modal deep learning algorithm. In detail, we will utilize transformers, a deep neural network architecture, that originates from the field of Natural Language Processing (NLP) and became popular in recent years in the field of computer vision. This method shows promising results in terms of understanding temporal and spatial information, which is essential in detecting and tracking highly dynamic eddies.

In this presentation, we introduce the deep neural network used in the EDDY project and show the results based on gridded data sets for the Gulf stream area for the period 2017 and first results of single-track eddy identification in the region.

It is clear that extreme tropical cyclone systems are increasing over the warm tropical oceans of the globe, which particularly constitutes a serious concern in the coastal locations with major population density areas. A few analyses have previously been conducted to explore the intensity with other hurricane-related parameters. However, little progress has been made to implement such characteristics in an automated fashion to automatically learn non-linear relationships between physical variables at different time windows. A novel hybrid framework based on supervised and unsupervised machine learning (ML) techniques is presented here to identify and optimally combine a mixed set of early predictors to forecast the evolution of the system, especially those systems attaining the highest categories. The size and brightness temperature of a tropical storm or hurricane are some of the key parameters from Earth Observation data that significantly contribute to better forecasting the intensity of a tropical cyclone system, including tropical storms, category 1-2 (minor hurricanes) and category 3-5 (major hurricanes) hurricane-strength systems. An 80%-20% train-test split was applied for evaluating our ML algorithms over the period from 1995 to 2019 in both the North Atlantic and East Pacific Oceans. Several metrics were also examined to assess robustness of our forecasting models, such as the accuracy rate (AR) or the Cohen’s Kappa value ( 𝜅 ). The hurricane intensity forecast accuracy is tied to the forecast window, being a 16- to 40-hour time window the most accurate one with about 70% AR and 𝜅 value of at least 0.4 (from moderate to perfect agreement). However, our models can also provide reasonable predictions of a hurricane’s intensity up to 56 hours ahead. Overall, a slightly higher intensity forecasting performance was found over the Atlantic Ocean. This promising framework is designed to include additional parameters and classes where relevant for further study.

Solar-Induced chlorophyll Fluorescence (SIF) is a crucial parameter for Earth Observation, as it is strictly related to the vegetation health status. In particular, SIF can be easily monitored through optical remote sensing and offers unique information about vegetation functional state. In fact, SIF emission is one of the three main pathways exploited by vegetation to convert the Absorbed Photosynthetic Active Radiation (APAR) and it is related to pivotal parameters such as the Gross Primary Productivity (GPP), which is a key-role parameter in many environmental applications.
In this contribution, we present a novel approach based on an optimized multi-parameter retrieval algorithm to improve the understanding of the complex relationships between fluorescence and biophysical/biochemical variables. The proposed algorithm is specifically aimed to analyse the reflectance spectra acquired from the vegetation and consistently retrieve the Top Of canopy SIF spectrum, the SIF spectrum corrected for leaf/canopy reabsorption (i.e. at photosystem level), the quantum efficiency (SIFqe) and three canopy-related biophysical parameters (Leaf Area Index - LAI, Chlorophyll content - Cab and APAR) in few milliseconds. This algorithm mainly consists on a novel hybrid Phasor-Machine Learning based approach, which is exploited for the first time for quantitative retrievals in the context of remote sensing studies, and it is the first method capable to retrieve the full fluorescence spectrum at the photosystem level and the fluorescence quantum yield from experimental measurements acquired onsite.
More in detail, our approach exploits reflectance spectra, which are discrete-Fourier transformed on consecutive spectrally resolved complex planes. In each considered complex plane, the spectra which are characterized by the same biophysical and SIF parameters are projected in the same point. It is therefore possible to predict the unknown properties of a single reflectance spectrum by evaluating its projection position in each plane. In order to exploit, for each estimated parameter, the most suitable spectral windows and at the same time, avoiding possible superposition effects in some planes, the algorithm employs a supervised Machine Learning algorithm, trained with the atmosphere-canopy radiative transfer (RT) SCOPE model, which analyses the projection position of the reflectance spectrum in all the considered planes and estimates the investigated variables.
The algorithm has been validated by means of RT simulations, characterizing its retrieving accuracy by varying different parameters (spectral windows width, number of exploited phasor planes, size of the dataset, etc.) and by applying to the analysed spectra an increasing Poissonian noise, described in terms of signal to noise ratio (SNR). Considering the experimental conditions (SNR >= 500), the algorithm is able to independently estimate each biophysical parameter and SIF spectrum with a relative root mean square error (RRMSE) lower than 5%.
In order to investigate the seasonal and daily dynamics of SIF, LAI, Cab, SIFqe and APAR, the method has been also applied to field experimental data collected in the context of the AtmoFLEX and FLEXSense ESA campaigns. Field data were acquired in Grosseto (Italy) from two different crops (forage and Alfa Alfa) by means of the FLOX spectrometer accommodated on ground at top-of-canopy level and on high towers (~100 meters) in a deciduous forest of Downy Oak (France).
The retrieved annual dynamic for SIF spectra has been then compared with the results obtained by state of the art inversion-based methods, showing a good consistency between the two different approaches (RRMSE ~ 10%). Moreover, also the daily dynamics of the investigated variables behave accordingly to the results of the theoretical models.
The retrieval of SIF at the high tower has been investigated excluding the O2 spectral bands affected by the atmospheric reabsorption. The obtained results are promising and it has implications on tower-based measurements, where complex and computationally expensive atmospheric compensation techniques are needed to retrieve fluorescence from oxygen absorptions bands.
In summary, in this study we will show the results of the performance of this new algorithm and its ability in deriving accurate fluorescence spectrum corrected for reabsorption and fluorescence quantum yield from model simulation and real measurements collected over two crops and on a deciduous forest. Overall, the obtained results are of undoubted use and exploitation for the ESA Earth Explorer FLEX Mission, and this study demonstrates a promising potential to exploit ground and tower spectral measurements with advanced processing algorithms, for improving our understanding on the link between canopy structure and physiological functioning of plants and it can be straightforwardly employed to process reflectance spectra to open new perspectives in fluorescence retrieval at different scales.

Complex numerical weather prediction (NWP) models are deployed operationally to predict the future state of the atmosphere. While these models solve numerically a system of partial differential equations based on physical laws, they are computationally very expensive. Recently, the potential of deep neural networks has been explored in a couple of scientific studies to generate bespoken weather forecasts for short time scales, inspired by the success of video frame prediction models in computer vision. In this presentation, we explore the application of different deep learning networks from the field of video prediction for weather forecasts up to 12 hours and discuss the potential and limitations of these approaches.

In the first study, we focus on the predictability of the diurnal cycle of near-surface temperatures. A ConvLSTM, and an advanced generative network, the Stochastic Adversarial Video Prediction (SAVP), are applied to forecast the 2 m temperature for the next 12 hours over Europe. Results show that SAVP is significantly superior to the ConvLSTM model in terms of several evaluation metrics. Our study also investigates the sensitivity to the input data in terms of selected predictors, domain size and number of training samples. The results demonstrate that additional predictors, i.e. in our case the total cloud cover and the 850 hPa temperature, enhance the forecast quality. The model can also benefit from a larger spatial domain. By contrast, the effect of reducing the training dataset length from 11 to 8 years is rather small. Furthermore, we reveal a small trade-off between the MSE and the spatial variability of the forecasts when tuning the weight of the L1-loss component in the SAVP model.

In the second study, we explore a custom-tailored GAN-based architecture for precipitation nowcasting. We developed a novel method named Convolutional Long-short term memory Generative Adversarial Network (CLGAN), to improve the nowcasting skills of heavy rain events with deep neural networks. The model constitutes a GAN architecture whose generator is built upon an u-shaped encoder-decoder network (U-Net) equipped with recurrent LSTM cells to capture spatio-temporal features. A comprehensive comparison between CLGAN, the advanced video prediction model PredRNN-v2 and the optical flow model DenseRotation is performed. We show that CLGAN outperforms in terms of point-by-point metrics as well as scores for dichotomous events and object-based diagnostics. The results encourage future work based on the proposed CLGAN architecture to further improve the accuracy of precipitation nowcasting systems.

In atmospheric remote sensing, the quantities of interest (e.g. the composition of the atmosphere) are usually not directly observable but can only be inferred indirectly via the measured spectra. To solve these inverse problems, retrieval algorithms are applied that usually depend on complex physical models, so-called radiative transfer models (RTMs). These are very accurate, however also computationally very expensive and therefore often not feasible in combination with the strict time requirements of operational processing. With the recent advances in machine learning, the methods of this field, in particular deep neural networks (DNNs), have become very interesting in order to accelerate and improve the classical remote sensing retrieval algorithms. However, their application is not straightforward as they can be used in different ways and there are many aspects to consider as well as parameters to be optimized in order to achieve satisfying results.

For the inverse problem in atmospheric remote sensing, there are two main approaches to apply neural networks:

1. Neural networks for solving the direct problem, where a neural network approximates the radiative transfer model and can thus replace it as a forward model for the inversion algorithm
2. Neural networks for solving the inverse problem, where a neural network is trained to infer the atmospheric parameters from the measurement (and some additional information, e.g. surface properties, viewing geometry) directly

For the first case, we present a general framework for replacing the RTM with a DNN that offers sufficient accuracy while at the same time increases the processing performance by several orders of magnitude. It's application is shown at the example of the ROCINN algorithm which is used for the operational cloud product of the copernicus satellites Sentinel-5 Precursor (S5P) and Sentinel-4 (S4).

The second case is more challenging: In contrast to approximating the radiative transfer model which provides a well defined function from the parameters to the measurements, the inverse problem is ill-posed. Due to this, small differences in the measurements can lead to large differences in the retrieved quantities. It is therefore desirable, to characterize the retrieved values by an estimate of uncertainty describing a range of values that are probable to produce the observed measurement. This can be achieved by using a bayesian framework, like it is done in bayesian neural networks or more novel methods like invertible neural networks. Based on these techniques, we show first results of the retrieval using neural networks for solving the inverse problem.

Finally, we will compare the two different approaches of using DNNs for the retrieval of cloud properties for S5P / S4 and discuss their applicability for the future.

Earth observation satellites are used for many tasks, among them the monitoring of agricultural areas. For example subsidence farming can be monitored to predict food shortages, which in turn helps organizing humanitarian aid faster when there is a shortage. Often, optical data is used because of the easy interpretability and especially the normalized difference vegetation index (NDVI) is frequently used for vegetation monitoring. However, in tropical areas with frequent cloud coverage or subtropical areas where the main growing season is during the rainy season clouds hinder the acquisition of optical images. To avoid this, active cloud-penetrating sensors like synthetic aperture radar (SAR) can be used. However, the greatly different characteristics of SAR images make interpretation more difficult and usable intelligence is harder to obtain.
There is great demand to mitigate this problem by converting SAR backscatter values to artificial NDVI values, which then in turn can be used for downstream tasks. This idea was already demonstrated in two studies [1, 2]. However, these studies are limited to small areas and present conversion models that are not globally applicable. Additionally, they suffer from a low performance when only relying on backscatter values and not using additionally data sources like the last cloud free NDVI value.
As solution we present a globally applicable model for the conversion of SAR backscatter values to NDVI values using a deep neural network. The used model does not rely on optical data at application time and is therefore unaffected by cloud cover.

To train the model, a dataset consisting of Sentinel-1 SAR data and Sentinel-2 optical data is created. To have a direct relation between backscatter and NDVI values the temporal distance is at most 12 hours between images of the same area. This avoids other influences like seasonal changes or vegetation growth. Images were sampled globally with an equal distribution for climate zones and land covers to capture the full spectrum of earth surfaces and vegetation. As auxiliary data, the 10m resolution ESA WorldCover product [3] and the 30m resolution ALOS JAXA DEM [4] were retrieved. Google Earth Engine was used to download the data.
The used model is a slightly adapted UNet. It does a pixel-wise regression of the NDVI using the VV and VH polarizations of the Sentinel-1 data, the ESA WorldCover and the ALOS DEM.

Using this approach, a globally applicable model is created to predict the NDVI from cloud-penetrating SAR images. This removes the need to train models for specific regions and vegetation. One disadvantage of this approach is the lower resolution of Sentinel-1 images (with a pixel size of 20x22m [5]) compared to the 10x10m resolution of Sentinel-2 images. This prevents the correct prediction of some fine spatial details. Further research is needed to increase the resolution regarding those details, either by using time series as input instead of images of a single date or by using other data sources to include more structural details.



ACKNOWLEDGMENT

This work was supported by the German Federal Ministry for Economic Affairs and Energy in the project “DESTSAM - Dense Satellite Time Series for Agricultural Monitoring” (FKZ 50EE2018A).


REFERENCES

[1] G. Scarpa, M. Gargiulo, A. Mazza, and R. Gaetano, “A CNN-based fusion method for feature extractionfrom sentinel data,” Remote Sensing, vol. 10, no. 2, Art no. 236, 2018.
[2] R. Filgueiras, E. C. Mantovani, D. Althoff, E. I. F. Filho, and F. F. da Cunha, “Crop NDVI monitoring based on Sentinel 1,” Remote Sensing, vol. 11, no. 12, Art no. 1441, 2019.
[3] D. Zanaga, R. Van De Kerchove, W. De Keersmaecker, N. Souverijns, C. Brockmann, R. Quast, J. Wevers, A. Grosu, A. Paccini, S. Vergnaud, O. Cartus, M. Santoro, S. Fritz, I. Georgieva, M. Lesiv, S. Carter, M. Herold, Linlin Li, N. E. Tsendbazar, F. Ramoino, O. Arino, ”ESA WorldCover 10 m 2020 v100,” 2021.
[4] J. Takaku, T. Tadono, M. Doutsu, F. Ohgushi, and H. Kai, “Updates of ‘AW3D30’ ALOS Global Digital Surface Model with Other Open Access Datasets”, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, ISPRS, vol. XLIII-B4-2020, pp. 183–189, 2020.
[5] Collecte Localisation Satellites, ”Sentinel-1 product definition,” Online, 2015.

The monitoring of the ice cover is important primarily for navigation, but also to study the water cycle and surface energy flux ; snow and ice cover on lakes is one of the 50 ECVs (Essential Climate Variables). The European Environment Agency (EEA) recently released near real time (NRT) snow and ice classification products over the EEA38 + UK European zone, using Sentinel-2 imagery as well as Sentinel-1 SAR data, with a decametric resolution : https://land.copernicus.eu/pan-european/biophysical-parameters/high-resolution-snow-and-ice-monitoring
The two main NRT products, derived from Sentinel-2 L2A MAJA images (https://logiciels.cnes.fr/fr/content/maja) are the Fractional Snow Cover (FSC) indicating the fraction of snow on 20m pixels, and the River and Lake Ice Extent (RLIE), also at 20m resolution, indicating the presence of ice / water / other on the river and lake Eu-Hydro mask. Barrou Dumont et al. (2021) show a very good match between in-situ data and the FSC products, and Kubicki et al. (2020) also show a good performance for RLIE products through comparisons with in-situ data as well as very high resolution SPOT 6/7 and Pléiades 1A/1B images. However, the RLIE validation results were obtained mainly on northern regions during winter, and were later shown to present a large number of ice classification false positives on turbid waters and salt lakes. This presents a problem both for climate analysis because of the large number of ice false positives during summer on turbid waters, and for the generalization of the algorithm to the full globe. Thin ice or black (transparent crystalline) ice also often goes undetected by the RLIE algorithm. Furthermore, the RLIE product requires the EU-Hydro water mask which is static and therefore limits the ability to track variability in river beds or lake surfaces, and misses a large number of small rivers and lakes not represented in the EU-Hydro water mask.
The RLIE algorithm relies on a minimum distance classifier as well as specific thresholds on certain bands to reduce false positives on turbid waters. We show that using more capable machine learning and deep learning methods, we are able to almost completely remove ice false positives on salt lakes and turbid waters, considerably improve thin ice and black ice detection, as well as generate products on the full images without the need for an a priori water mask.
Machine learning and deep learning results were obtained on the basis of 32 fully labelled Sentinel-2 images covering different regions on the globe, representing various cases including ice melt, ice formation over river and lakes, salt lakes, urban areas, fields, forests, mountain regions, as well as various turbid water cases.
For the machine learning approach, pixels are classified using solely the spectral information on the different Sentinel-2 L2A bands as well as various normalized difference indexes (NDSI, NDVI, NDWI, ...). Both the linear SVM and the RandomForest methods were evaluated : SVM yields better results as RandomForest has a tendency to overfit the input data and produce noisier results. The classification using those methods is considerably improved when compared to the RLIE product. Ice false positives on salt lakes and turbid waters are almost completely removed, but there are still some ice false positives, mostly isolated pixels, which yields somewhat noisy results in some cases. Thin ice / black ice classification is also much improved when compared to RLIE but in some cases the ice still goes undetected although a visual inspection of the image by a human operator clearly recognises the ice. This is due to our ability to use spatial pattern recognition and exploit water / ice borders as well as cracks in the ice.
To improve on the machine learning classification, we used deep learning to exploit spatial information. Two neural networks were evaluated: EfficientNet + DeepLabV3+ as well as EfficientNet + RefineNet. Preliminary results at the date of this abstract indicate that the deep learning approach effectively reduces noise in the classification of products, completely removes ice false positives over turbid waters and adequately classifies ice that was missed in the machine learning approach despite having very visible cracks and water/ice borders. However, in many cases, the borders between categories are less accurate than using the machine learning approach, and manual setting of weights in the cost function is necessary to fine tune the algorithm and avoid over or under classification of the less represented categories in the database (salt lakes are the less represented). Those results are however preliminary and will be consolidated in the coming months.

Barrou Dumont, Z., Gascoin, S., Hagolle, O., Ablain, M., Jugier, R., Salgues, G., Marti, F., Dupuis, A., Dumont, M., and Morin, S.: Brief communication: Evaluation of the snow cover detection in the Copernicus High Resolution Snow & Ice Monitoring Service, The Cryosphere, 15, 4975–4980, https://doi.org/10.5194/tc-15-4975-2021, 2021.

Kubicki, M., Bijak, W., Banaszek, M., Jasiak, P., High-Resolution Snow & Ice Monitoring of the Copernicus Land Monitoring Service Quality Assessment Report for Sentinel-2 ice products, 2020 : https://land.copernicus.eu/user-corner/technical-library/hrsi-ice-qar

There is no doubt that the devastating socio-economic impacts of floods have increased during the last decades. According to the International Disaster Database (EM-DAT), floods represent the most frequent and most impacting event among the weather-related disasters regarding the number of people affected. Nearly 1 billion people were affected by inundations in the decade 2006–2015, while the overall economic damage is estimated to be more than $300 billion, with individual extreme events, like Superstorm Sandy, costing several tens of billions in damage (an estimated $65 billion in damages in the US). Despite this evidence and the awareness of the environmental role of rivers and their inundation, our capability to respond to and forecast floods remains very poor, mainly due to the lack of measurements and ancillary data at large scales.
In this context, satellite sensors represent a precious source of observation data that could fill many of the gaps at the global level, especially in remote areas and developing countries. With the proliferation of more satellite data and the advent of ESA's operational Sentinel missions under the European Commission's Copernicus Programme, satellite images, particularly SAR, have been assisting flood disaster mitigation, response, and recovery operations globally. In addition, the proliferation of open satellite data has advanced the integration of remotely sensed variables with flood modeling, which promises to improve our process understanding and forecasting considerably.
In recent years, the scientific community has shown how earth observation can play a crucial role in calibrating and validating hydraulic models and providing flood mapping and monitoring applications to assist humanitarian disaster response.
Although the number of state-of-the-art and innovative research studies in those areas is increasing, the full potential of remotely sensed data to enhance flood mapping has yet to be unlocked, especially since the latency issue is not being sufficiently well addressed. Indeed, the time between image acquisition to the flood map delivery to the person who needs it is not in line with disaster response requirements. For instance, at the moment, almost all flood maps reach the field teams after two days of image acquisition, which renders the map pretty much useless for instant rescue operations. A delay of 3 days renders the map unusable for almost all operation stages. While until recently delays appeared unavoidable due to the mapping process not being highly automated by transferrable AI, the lion's share of time loss has now shifted to the communication of requests and data. Indeed, it is now responsible for the slow uptake of EO-based products, such as flood maps, into an operational timeline or disaster response protocol of various potential user organizations, such as the UN World Food Programme.
In this work, we present the conception of a Digital Twin Experiment (DTX) to generate a prototype AI-based solution that could be deployed onboard SAR satellites to produce flood maps in near-real-time.
With our contribution, we aim to conceptualize the processing needed onboard the satellite: from the SAR image formation to the Machine Learning-based flood detector and classifier. On the other hand, we provide a strategy for the fast delivery of the onboard inference to achieve short latency. The mapping results consist of column/row raster vectors indicating the flooded area. Therefore, after the creation of a flood map on the "orbital edge", it will be sent in the form of a short "message" and delivered to the field response teams via satellite communication technology for use within minutes, rather than many hours to days as it is currently the case.
We aim to simulate various scenarios by varying the quality of the input SAR data and experimenting with Machine Learning approaches for image segmentation. We ground our experiments on three different viewpoints. First, we aim to assess the impact of SAR image quality on the final result. Therefore, we simulate with DLR's End-2-End SAR simulator different scenes at different resolution and noise levels. Pre- and post-flood scenarios will be considered. Secondly, we also consider possible approximations in the SAR focusing.
On the one hand, we could suppose that it will be possible to obtain a perfectly focused SAR image onboard and with fast computational time. On the other hand, we could already assess the performance of a non-perfectly focused SAR image on the final classification result. For this purpose, we experiment by approximating the standard focusing kernel of a strip-map acquisition mode. We provide, therefore, a dataset of SAR images with different deformations due to the non-perfect SAR focusing. An example of focusing kernel approximation might be to perform unfocused azimuth processing. As a final aspect, we will consider different deep learning models for semantic segmentation, trained to perform at their best in the different proposed scenarios. Eventually, the performance comparison will provide the best trade-off between achieved classification performance and computational effort of the entire processing chain, including SAR focusing. We eventually propose a holistic evaluation strategy for the proposed end-to-end framework, which will provide highly representative and practically oriented metrics for mapping accuracy and processing efficiency.

Identifying regions that have high likelihood for wildfires is a key component of land and forestry management and disaster preparedness. In recent years there has been considerable interest on wildfire modeling and prediction in the domain of machine/deep learning. Focusing on next day fire risk prediction, we have already conducted a thorough research on this specific problem [1, 2] achieving very promising results and gaining valuable insights into the complexities and specificities of task. In our previous work, we formalized the problem as a binary classification task between the fire/no-fire classes, considering as instances the daily snapshots of a grid with 500m-wide cells. Each instance is represented by earth observational, meteorological and topographical data features derived up until the previous day of the prediction. To this end, we utilize a massive dataset, labeled with fire/no-fire information for each grid cell and in daily granularity, covering the whole Greek territory for the years 2010-2020 (focusing mainly on months April to October, which correspond to the main fire season). In this paper, we discuss the major specificities of the task that we have identified by our work so far and propose a concrete Deep Learning (DL) framework that has the potential to jointly handle them.
0.1 Next day fire prediction specificities
Highly imbalanced dataset. The amount of instances that represent the classes of fire and no-fire follow a different distribution with the respective ratio being of the order ∼1 : 100,000. As a conse- quence, vanilla Machine Learning (ML) algorithms are expected to learn biased classification models, due to their inherent design to optimize based on accuracy-like measures, which are improper for imbalanced datasets [3]. As a consequence, the most important class (representing risk of fire occurrence in our case) is poorly modeled and predicted. Various techniques can be employed to mitigate this issue, including class over/under-sampling and cost-sensitive training (i.e. assigning different
weights to the cost of instances of different classes, during the algorithms training/optimization process) [3]. Although these techniques have been extensively studied in the past [4], applying them as is, without adaptation to the task’s particularities, can become problematic. For example, random under-sampling could exclude valuable information from the dataset, whereas oversampling may introduce undesirable noise to the data by interpolating new points between marginal outliers and inliers (data observations that lie in the interior of a data set class in error). Moreover, these approaches do not eventually model the true data distribution of the problem, thus present limited capability to solve real world problems. Finally, such techniques are methodologically poorly applied, e.g. by also performing over-/under-sampling also on the test setting, resulting to poorly, even erroneously, assessed techniques. Pseudo-negative class (alt. absence of fire). Adding to the above problem of fire instances sparseness, there exists a considerable amount of no-fire instances that are not really negative examples, but in fact they denote the absence of fire. This can be misleading since it does not necessarily mean low fire risk, but lack of fire precursor for a wildfire to start. Hence, these particular instances lie very close to fire instances in the feature space, which makes it harder for the models to be properly trained, since the decision boundaries between these cases cannot be clearly established. Thus, traditional ML models demonstrate poor class separability, mainly in the samples lying close to the decision boundaries. We performed an indicative similarity analysis between the fire/no-fire instances and between the fire instances in subsets of our dataset comprising separate months. The results for
August 2010 are reported in the Tables 1 and 2, where similarity class 1 denotes dissimilar instances and similarity class 10 means very similar instances. The following findings are pointed out:
•More than 70% of the fire instances are very similar (similarity class>= 6) to no-fire instances, while
•The similarity between the fire instances is split among the classes 5 to 9 with class 9 having the highest percentage as expected. It also observed that some fire samples are perfectly aligned (class 10), which is explained by the high spatiotemporal correlations inherit in the data [2].

The similarity measure is based on the calculation of the Euclidean distance between the normalized feature vectors of the instances, and was performed for the 1K most similar samples on a monthly basis. The demonstrated results are representative and referred to August of 2010, as similar patterns were found to all the analyzed months and years.

Concept Drift. Another particularity comprises concept drifts (alt. data shifts) that were detected in some feature cases, between the different months and years. For example, the meteorological parameters fluctuate periodically during the year, but also get altered through the years due to the climate change. A typical drift example of data is presented in Figure 1, where a significant drift of the mean temperature, for both fire and no-fire classes, is observed in the year 2012. Further, this feature has strong fluctuations mainly in the fire class through the years. It also worth noticing that in 2018, the mean temperature of the no-fire cells exceeds the mean temperature of the fire cells. Additionally, although at most years fire and no-fire boxplots present significant range differences (i.e. 2010 and 2013), in 2016 the variation of fire samples is a subset of the no-fires’ dispersion, which makes class separability impossible. These underlying changes in the statistical properties of the
data, could degrade the predictive performance of the ML models.
0.2 Handling next day fire prediction via Siamese Networks
We consider the deployment and adaptation of supervised deep metric learning architectures, like Siamese Neural Networks (SNN) [5], as a promising framework for handling the aforementioned specificities of the next day fire prediction task. A SNN consists of two identical NN architectures which are trained in parallel. One sample is passed to the first network, an other to the second and the system is trained on a distance function between the vector representations produced by the identical networks. The aim is to generate good representations of the data, in order to bring similar samples closer and dissimilar samples further away in the distributed representation space. Various loss functions can be selected for optimization, including triplet loss functions [6], which we believe that have great potential for the specific problem examined in this paper. At each iteration of the training process a baseline (anchor) instance is compared to a positive (same class) and a negative (opposite class) input, and the system tries to minimize the distance between the anchor-positive samples and maximize the distance between anchor-negative samples. This training schema is adjustable to our data and could undertake several modifications, extensions and customizations in order to deal with several of the specificities of the problem. Firstly, in order to deal with the imbalanced dataset and the absence of fire concept, the loss function could be customized to generate the triplets by filtering difficult examples from the majority class instead of randomly select the positive and negative samples. Another approach would consist in the relabeling of the noisy samples of the no-fire class to fire class ones. These actions would not only redefine the class boundaries during training, in order to facilitate the identification of more meaningful class boundaries, but also ameliorate to some extent the extreme class imbalance. Considering concept drift, we have the intuitions that distinct cases of the fire class (rare cases [3]) could be identified and split into distinct classes, corresponding to clusters of fires with different characteristics. Properly adjusting the triplet generation process to create triplets dedicated for each of the different fire classes, could be an additional tool towards learning better (and more specific) representations for the fire instances and, eventually, more accurately handling the problem.
References
[1] Alexis Apostolakis, Stella Girtsou, Charalampos Kontoes, Ioannis Papoutsis, and Michalis Tsoutsos. Implementation of a random forest classifier to examine wildfire predictive modelling in greece using diachronically collected fire occurrence and fire mapping data. In Jakub Lokoc, Tom ́as Skopal, Klaus Schoeffmann, Vasileios Mezaris, Xirong Li, Stefanos Vrochidis, and Ioannis Patras, editors, MultiMedia Modeling - 27th International Conference, MMM 2021, Prague, Czech Republic, June 22-24, 2021, Proceedings, Part II, volume 12573 of Lecture Notes in Computer Science, pages 318–329. Springer, 2021.
[2] Stella Girtsou, Alexis Apostolakis, Giorgos Giannopoulos, and Charalampos Kontoes. A machine learning methodology for next day wildfire prediction. In IGARSS, 2021.
[3] Haibo He and Yunqian Ma. Imbalanced Learning: Foundations, Algorithms, and Applications. Wiley-IEEE Press, 1st edition, 2013.
[4] Qiang Yang and Xindong Wu. 10 challenging problems in data mining research. International Journal of Information Technology Decision Making, 05(04):597–604, Dec 2006.
[5] Vijay Kumar B. G, Gustavo Carneiro, and Ian Reid. Learning local image descriptors with deep siamese and triplet convolutional networks by minimising global loss functions. arXiv:1512.09272 [cs], Aug 2016. arXiv:1512.09272.
[6] Gal Chechik, Varun Sharma, Uri Shalit, and Samy Bengio. Large Scale Online Learning of Image Similarity through Ranking, volume 5524 of Lecture Notes in Computer Science, page 11–14. Springer Berlin Heidelberg, 2009.

In recent years we witnessed an increasing number of remote sensing data being available through public satellite data sources (particularly Copernicus Sentinel and Landsat) and distributed in public cloud infrastructures like Amazon Web Services, Google Cloud and the various DIAS and national mirroring infrastructures throughout Europe. Also this evolution overlaps with the recent technological developments as is the introduction of Cloud Enabled GeoTiff, facilitating easy "data cube" like access to exposed data and also the availability of tools like Stackstac enabling friendly access from common data processing tools like Dask.

Both the available data and tools represent a great opportunity for employing Machine Learning methods on a wider scale, integrating various types of satellite and in-situ observations and opening the door towards new techniques and thematic fields.

In this context, we present an updated version of the Hugin EO Machine Learning tool providing support for STAC (Spatio Temporal Asset Catalog) based data cubes and for emerging technologies like Zarr, XArray and stackstac, providing Cloud Native data access.

One of the advantages of HuginEO is it's interoperability with Jupyter based notebooks, enabling users to easily experiment with new models, visualize predictions and analyze various metrics.

Also, we introduce support for additional backend Machine Learning technologies and model types (eg. self-supervised models), and extended support for hyper-parameter model optimization. All models are tested and evaluated against a number of use cases referring to the exploitation of spatio-temporal and spectral features of satellite data, considering their expected use in agriculture and forestry for monitoring purposes and possible integrations in national, thematic, user-oriented Earth Observation data platforms. The current evaluation of the model performance, considering the state of the art metrics, show promising results.

HuginEO is accompanied with a suite of pretrained models for crop classification, building detection in VHR data, super-resolution and models trained using self-supervised techniques (usable for transfer learning in various other applications).

Satellite Earth Observation missions and data analysis are critical elements of each segment of the Earth Observation value chain. New datasets and observations lead to new knowledge of physical, chemical, and biological processes of the Earth system. When used to improve Canadian Earth System models, environmental prediction and climate projection services to Canadians are made better, more accurate and robust. For over two decades, the Canadian Space Agency (CSA) has funded EO missions and scientific research to advance satellite and instrument operations, data product development and validation, and data analysis. This presentation will focus on current CSA supported missions, their new data products and validation, recent scientific advances they have enabled, along with new research results from satellite data analysis projects. The talk will also highlight academic-government collaborations, the Earth system models they advance, and international collaborations enabled.

Landsat satellites have been providing continuous monitoring of the Earth’s surface since 1972. The US Geological Survey (USGS) and the National Aeronautics and Space Administration (NASA) entered into an interagency agreement for Sustainable Land Imaging (SLI) to continue the Landsat quality global survey missions and 50 years of Landsat data record. Landsat 9 is the first SLI mission, which was successfully launched on September 27th, 2021 from the Vandenberg Space Force Base. Landsat 8 and 9 together will provide the best quality Landsat observations yet from space to support food security, monitor water use, assess wildfire impacts and recovery, monitor forest health, track urban growth, support climate resiliency and grow the economy. Planning for the Landsat 9 follow-on mission, Landsat Next, is already underway. The USGS National Land Imaging Program has collected land imaging user needs from a range of applications across the Federal civil community and other stakeholders to help define Landsat Next science objectives and requirements. User community has expressed great interests in maintaining Landsat continuity, supporting synergy with Copernicus Sentinel-2 mission and enabling new emerging applications that are critical to tackle the challenges in today’s global environment. Landsat Next draft science requirements have included improvements in spatial resolution, temporal revisit and spectral capability while maintaining science data quality to continue serving the global land imaging community. This presentation will provide an overview of the SLI user needs process, Landsat Next draft science requirements and Landsat Next mission status.

The NASA-ISRO Synthetic Aperture Radar (NISAR) mission will use synthetic aperture radar to map Earth’s surface every 12 days, persistently on ascending and descending portions of the orbit, over all land and ice-covered surfaces. The mission’s primary objectives will be to study Earth land and ice deformation, and ecosystems, in areas of common interest to the US and Indian science communities. This single observatory solution with an L-band (24 cm wavelength) and S-band (10 cm wavelength) radar has a swath of over 240 km at fine resolution, and will operate primarily in a dual-polarimetric mode in an exact repeat orbit.

NISAR will characterize long-term and local surface deformation on active faults, volcanoes, potential and extant landslides, subsidence and uplift associated with changes in aquifers and subsurface hydrocarbon reservoirs, and other deforming surfaces. These measurements will be used to model the physics of the subsurface, potential hazards associated with the deformation, and associated risks. The variable and largely unpredictable nature of these phenomena lead to a systematic collection strategy to capture as many signals as possible. Surface deformation measurements will be validated over globally distributed GPS networks in a variety of environmental settings.

NISAR will determine changes in carbon storage and uptake resulting from disturbance and subsequent regrowth of global woody vegetation, by regularly measuring the amount of woody biomass and its change in the most dynamic regions of the world. NISAR also has objectives in characterizing changes in the extent of active crops to aid in crop assessments and forecasting, as well as changes in wetlands extent, freeze/thaw state and permafrost degradation. The ecosystems data sets will be validated through in situ measurements at dozens of sites around the world in partnership with other missions and organizations.

NISAR will investigate the nature and causes of changes to Earth’s ice sheets and sea ice cover in relation to the atmospheric and ocean forces that act upon them, through systematic deformation measurements of Greenland’s and Antarctica’s ice sheets, seasonal dynamics of highly mobile and variable sea ice, and inventory the variability of key mountain glaciers which are retreating in many places at a record pace. Validation of deformation on the ice sheets will employ bare-rock references and cross-over analysis, as well as some deployed GPS stations on the flowing ice. The sea ice community will compare sea ice motion to in situ buoy data.

In addition, NISAR will be operated to observe potential hazards and disasters on a best-efforts basis to demonstrate rapid assessments in urgent events such as earthquakes, volcanic eruptions, floods, and severe storms. These data will support research into effective rescue and recovery activities, system integrity, lifelines, levee stability, urban infrastructure, and environment quality. The mission team has implemented an urgent-response tasking system that will combine automated triggers for earthquakes, volcanoes, and fires with manual requests by certified users.

The joint science team at NASA and ISRO has created a stable, joint science and observation plan, robust calibration and validation plan. The team also has defined a suite of science products, including raw data, complex images at full resolution in both natural radar coordinates and in an orthorectified form, and lower resolution polarimetric and interferometric products also in radar and ortho-rectified coordinates. These products will be organized in frames roughly 240 km x 240 km in size, and will be available at the Alaska Satellite Facility Distributed Active Archive Center under NASA’s full and open data policy.

The science team has developed the observation plan prioritizing continuity of time series measurements. To that end, the polar measurements prioritize the South Pole, creating a coverage gap north of 77.5 degrees of latitude, due to the inclined orbit and consistent southward off-nadir pointing of the radars. The radar instruments have many possible modes, operated at L-band globally, and jointly with S-band regionally over India and select other locations around the world. The NISAR science observation plan is designed to tackle the science questions posed by persistent and consistent imaging of Earth’s land and ice surfaces throughout the life of the mission, delivering time series of approximately 30 images per year from ascending and descending vantage points.

The cadence of science operations is expected to be highly routine. The initial observation plan will be in place pre-launch, and it is anticipated that there will only be minor adjustments to the plan once in orbit. The project has defined a six-month replanning cycle, whereby scientists identify changes they would like to see based on the data previously acquired and analyzed science data, the science and project mission planning teams evaluate the impact of those changes on resources and the ability to meet science requirements, and if acceptable, the observation plan is revised. Each of these steps is allocated roughly 2 months. Given that the goal of NISAR is to create regular, easy-to-use, time series of Earth change, the project expects that any adopted changes will not break the time series.

The science team is developing algorithms to produce higher level products for validation purposes. These products will be over local or regional validation sites, with sufficient to demonstrate that the required accuracies can be achieved over the Earth. The algorithms for producing these products will be made available to scientists interested in using NISAR data. They are being developed in the form of jupyter notebooks, serving the purpose of processing, algorithmic description and documentation, and instruction. Sample data sets will be available prior to launch to prepare the community for the data products and the algorithmic workflows to produce higher level products.

NISAR is in its third phase of integration and test, when all the components of the instrument payload, including the L- and S-band radar electronics, the solid-state recorder, GPS, and engineering payload, and the mechanical boom and reflector systems are assembled and tested in environments at NASA’s Jet Propulsion Laboratory. This completed payload will be subsequently shipped to India for the final phase of integration, test and launch, currently planned for 2023. The mission systems have been built and are in extended operational testing.

With NOAA’s current Geostationary Operational Environmental Satellites – R Series (GOES-R) satellites slated to end their operational service in the mid-2030s, attention has turned to planning NOAA’s next generation system, the Geostationary Extended Observations (GeoXO) series.

Pre-formulation activities for GeoXO were conducted over 2020-2021 and included a wide-ranging assessment of user needs via workshops, conferences, outreach events, and surveys; the evaluation and prioritization of potential observational choices versus NOAA mission service areas; documentation of the societal benefits for each observation; industry studies of instruments and architecture concepts; and government-led instrument and constellation studies. The pre-formulation phase resulted in the definition of observational requirements for this new system. GeoXO will continue the GOES-R-legacy observations of visible/infrared imagery and lightning mapping that are critical for tracking real-time environmental conditions. In addition, GeoXO will provide new observations to improve weather forecasting including hyperspectral sounding, to aid numerical weather prediction and nowcasting, and potentially low light imagery, for tracking clouds and smoke at night. GeoXO will also meet new needs for ocean, coast, and atmospheric monitoring with ocean color imagery and atmospheric composition sensing. Architecture trades were conducted in order to select a GeoXO constellation and the program completed a Mission Concept Review in June 2021.

The planned GeoXO constellation will include twin “east” and “west” satellites with the Imager, Lightning Mapper, and Ocean Color instruments, and a third “center” satellite carrying the Sounder and Atmospheric Composition instrument. The program was officially approved to begin formulation activities in November 2021. Phase A industry studies for the Imager and Sounder instruments are underway and studies for the other instruments and spacecraft are planned to begin in 2022. The first GeoXO launch is targeted for 2032 and the satellites are expected to be operational into the 2050s.

This presentation will discuss the GeoXO program scope, requirements, mission architecture, status, and timeline, along with the plans for program formulation, slated to continue through 2025.

A set of observing system simulation experiments (OSSEs) was preformed to investigate the impact of assimilating geostationary hyperspectral infrared sounder and microwave observations into the Global Modeling and Assimilation Office (GMAO) OSSE system. OSSEs and several other tools in this work is intended to help inform NOAA’s Geostationary Extended Observations (GeoXO) program to assess the potential gains from various configurations of geostationary infrared (IR) and microwave sounders from a numerical weather prediction perspective using a global system. Infrared sounder configurations consider systems with longwave thermal infrared-only and short-to-midwave infrared-only spectral coverage. Scenarios with two sounders at 75 degrees and 135 degrees West longitude and one sounder at 105-degree West longitude are also assessed along with other IR GEO sounders positioned around the Earth. The microwave sounder simulation is run in a similar configuration as the experiment with two infrared sounders, but several channels in the 60, 165, and 183 GHz frequency regions are used instead. A summary of progress and results will be presented.

The PACE mission represents NASA’s next great investment in ocean biology, clouds, and aerosol data records to enable advanced insight into ocean and atmospheric responses to Earth’s changing climate. Scheduled for launch in January 2024, PACE will not only extend key heritage essential climate variable time-series, but also enable the accurate estimation of a wide range of novel ocean, land, and atmosphere geophysical variables. A key aspect of PACE is its inclusion of an advanced hyperspectral scanning radiometer known as the Ocean Color Instrument (OCI) to measure the “colors” of the ocean, land, and atmosphere. Whereas heritage instruments observe roughly five to ten visible wavelengths from blue to red, OCI will collect a continuum of colors that span the visible rainbow from the ultraviolet to near infrared and beyond. Specifically, OCI is a scanning spectrometer that spans the ultraviolet to near-infrared region in 2.5 nm steps and also includes seven discrete shortwave infrared bands from 940 to 2260 nm, all with 1 km2 nadir ground sample distances and 1-2 day global coverage. This leap in technology will enable improved understanding of aquatic ecosystems and biogeochemistry, as well as provide new information on phytoplankton community composition and improved detection of algal blooms. OCI will also continue and advance many atmospheric aerosol, cloud, and land capabilities from heritage satellite instrumentation, which in combination with its ocean measurements, will enable improved assessment of atmospheric and terrestrial impacts on ocean biology and chemistry. The PACE payload will be complemented by two small multi-angle polarimeters (MAP) with spectral ranges that span the visible to near infrared spectral region, both of which will significantly improve aerosol and hydrosol characterizations and provide opportunities for novel ocean color atmospheric correction. The first MAP, the University of Maryland Baltimore County HARP2 instrument, will provide wide-swath multispectral polarimetric retrievals at 10-60 view angles. The second MAP, the SRON Netherlands Institute for Space Research and Airbus Defence and Space Netherlands SPEXone instrument, will provide narrow-swath hyperspectral polarimetric retrievals at 5 view angles. Ultimately, the PACE instrument suite will revolutionize studies of global biogeochemistry, carbon cycles, and hydrosols / aerosols in the ocean-atmosphere system and, in general, shed new light on our colorful home planet. This presentation will showcase the current status of the PACE mission, with a focus on instrument characteristics, core and advanced data products and their access, community engagement and potentials for new synergies and collaborations, and other mission plans as PACE heads towards its launch.

We developed a ResNet methodology based on Convolutional Neural Network (Zanchetta and Zecchetto, 2021), able to estimate the wind direction at a spatial resolution of 500 m by 500 m without external information. The ResNet model can derive the wind field even in the absence of wind streaks, in presence of convective turbulence structures, atmospheric lee waves, and ships. It is indicated to extract wind information over small areas, as the example of Venice lagoon. In this work the wind fields have been producet using the directions from ResNet and the scatterometer-based Geophysical Model Function CMOD7 (Stoffelen et al., 2017).
The possibility offered by ResNet led us to investigate the characteristics of the strongest winds blowing on the northern Adriatic Sea and Venice Lagoon, Italy. The area of interest is subjected to high spatial and temporal variability of wind, a peculiarity of many coastal areas, making it a very demanding site.
The structure of the wind systems inside and outside the lagoon has been studied in terms of spatial variability of speed, direction and vertical velocity wek in the Ekman layer derived by ResNet.
The layout of wek exhibits contiguous cells of upward and downward motion elongated orthogonally to the wind direction with periodicity of 5.4 km. This spatial variability seems to be a signature of the atmospheric Ekman pumping, produced by local variations of direction and speed.
An example of results from ResNet and OCN is reported in Fig. 1, which shows the ResNet (left panel) and the OCN SAR wind fields over the Venice lagoon: differently from the ESA OCN winds, the unprecedented resolution obtained with the ResNet allows an exhaustive coverage of the Venice lagoon, making possible to investigate the spatial structure of wind fields. For instance, under northeastern storms (Bora), the wind speed increases from northern to southern lagoon by 30% in average, in agreement with a case study carried out on experimental data (Zecchetto et al., 1997).
SAR winds derived by ResNet have been compared with the in-situ and ECMWF model data, showing on average, a 9% of underestimation and 7% of overestimation respectively, in the range from 4 ms-1 to 25 ms-1. The overestimation of SAR derived winds with respect to ECMWF confirms the results obtained in the Adriatic basin from comparisons between scatterometer and ECMWF winds (Zecchetto et al., 2015), while the underestimation with respect to the in-situ data conveys the difference between ECMWF and in-situ winds of ~10%.
The importance of a correct determination of the wind direction has been tested by comparing the SAR wind fields produced using ResNet and ECMWF wind directions, which may differ locally up to ±30º: these discrepancies may produce local differences of wind speed as large as ±2 ms-1.
Detailed analysis of selected cases raised the issue of the lack of data with true spatial resolution of O(1) km and within half hour from the satellite pass time necessary for exhaustive comparisons.


References

Stoffelen, A., Verspeek, A., Vogelzang, J., Verhoef, A., 2017. The CMOD7 Geophysical Model Function for ASCAT and ERS Wind Retrievals. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 10, 2123–2134, doi:10.1109/JSTARS.2017.2681806 .

Zecchetto, S. , G. Umgiesser and M. Brocchini, Hindcast of a Storm Surge Induced by Local Real Wind Fields in the Venice Lagoon, Continental Shelf Research, Vol.17 No.12,1513-1538, 1997

Zecchetto, S., della Valle, A., De Biasio, F., 2015. Mitigation of ECMWF–scatterometer wind biases in view of storm surge applications in the Adriatic Sea. Adv. Space Research 55, 1291–1299. doi:10.1016/j.asr.2014.12.011 .

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)

The Chinese French Ocean Satellite (CFOSAT) is an innovative space mission dedicated to the global observation and monitoring of the ocean sea state and the sea surface vector winds. CFOSAT operates two Ku-band rotating radars: the nadir/near-nadir Ku-band wave scatterometer (SWIM) and the dual-polarization, moderate incidence angle, Ku-band wind scatterometer (SCAT). This unique instrumental configuration provides regular collocated measurements of radar backscatter to retrieve sea surface state parameters, including significant wave height, directional wave spectrum, and wind vector. Two sensors also give the opportunity to improve the quality of the retrieved parameters by combining both data sources. In particular, this approach can be applied for the improvement of SCAT wind retrievals using SWIM observations.

The effective backscattering properties of SWIM and SCAT do not perfectly match the commonly used Ku-band Geophysical Model Function (GMF) due to various reasons like radar antenna design, swath patterns and noise signal distortions. On the other hand, observations for different incidence angles have different sensibility to sea surface parameters: short and long waves, surface currents, surface temperature, etc. The joint use of multi-instrument measurements within a common processing framework, i.e. scatterometer wind vector inversion Maximum Likelihood Estimator procedure, brings a potential risk of a significant error multiplication due to models and observation data mismatch.

The relation between collocated backscatter (σ0) measurements and various environmental parameters could be justified with the new common GMF which describes geophysical and CFOSAT-specific instrumental properties for all onboard sensors in a unified form. Such alternative Ku-band GMF was developed using a neural network (NN) approach. The traditional set of GMF variables (wind vector, incidence angle, polarization, ….) was extended with various additional geophysical parameters which can impact the signal properties: significant wave height, sea surface current vector, sea surface temperature, ice concentration, precipitation rate. The NN learning data set is based on CFOSAT measurements and collocated model data as provided by IFREMER Wave and Wind Operational Center (IWWOC) with SWISCA S Level 2 product. To avoid model biasing, special attention was addressed to the normalization and uniformization of input values during the learning process. As well, the numerical learning strategy was adapted to reduce the negative impact of using numerical weather prediction models (NWP) in the backscatter measurements regression task. The derived NN GMF reproduces the main features of NSCAT-4 GMF for moderate incidence angles and TRMM/GPM GMF for near-nadir observations. However, instrument-specific features are clearly present as well.

The resulting NN GMF could be considered as the approximation of Ku-band radar cross-section as a function of a multi-parameter environment. This function allows separating the impact of different geophysical variables on the backscattering coefficient value. The flexible nature of the proposed approach naturally enables the inclusion of any additional sea state variables to GMF. Additionally, it provides a robust platform for rapid signal calibration and re-adjustment during mission exploitation. We anticipate the implementation of the demonstrated model to extend the existing SCAT data processing with collocated SWIM nadir/near-nadir observations and additional NWP variables. As well, this approach can be suggested for the implementation in other scatterometry processing chains associated with different instruments and sensing microwave bands.

Radiance measurements from spaceborne microwave instruments are the most impactful observations used in Numerical Weather Prediction (e.g. Eyre, English and Forsythe 2020). Sophisticated data assimilation methods such as 4D-Var have been critical to this success, enabling direct assimilation of raw radiances. However, until recently, different Earth System components such as ocean, land and atmosphere were always handled separately, meaning those radiances which are sensitive to more than one component are still assimilated sub-optimally. The development of coupled data assimilation methodologies enables us to take another big step in the use of radiances, simultaneously and consistently fitting the state in multiple sub-systems to the same observations. This requires improved surface radiative transfer models.

For the ocean, although in general physically based models are used in data assimilation, at least for modelling passive microwave observations, the uncertainty is not well known and often different models are used for different spectral bands, and for active and passive sensing instruments. Furthermore, for active-instrument, empirical Geophysical Model Functions are used, which are very accurate (~0.1 dB), but physically-based methods have lower accuracy (Fois, 2015). In attempting error budget closure, lack of knowledge of uncertainty in surface emission was a limiting factor (GAIA-CLIM: www.gaia-clim.eu/). An International Space Science Institute team was created (English et al. 2020) to address this gap, taking the best available model components, integrating, testing across all spectral bands and characterizing as far as possible the uncertainty. The resulting reference model will then be provided as community software on GitHub.

In this short presentation, the choices made assembling this model will be explained, building on the starting point of the LOCEAN model of Dinnat et al. (2003). Samples of characterization undertaken will also be summarized. This includes comparison to SMAP, AMSR2 and GMI (e.g. Kilic et al. 2019) and early work to evaluate in the infrared and for active sensors. Finally, the plans for making code available will be briefly presented. This model will also be used to generate training data for fast models, e.g. Fastem (English and Hewison 1998), as used in operational data assimilation and climate re-analysis.

References

Dinnat, E. P., Boutin, J., Caudal, G., and Etcheto, J., 2003 : Issues concerning the sea emissivity modeling at L band for retrieving surface salinity, Radio Sci., 38, 8060, https://doi.org/10.1029/2002RS002637

English, S., Prigent, C., et al., 2020: Reference-quality emission and backscatter modeling for the Ocean, B. American Meteorol. Soc., 101(10), 1593-1601. https://doi.org/10.1175/BAMS-D-20-0085.1

English S.J. and Hewison T.J., 1998: Fast generic millimeter-wave emissivity model, Proc. SPIE 3503, Microwave Rem. Sens. Atmos. Env., https://doi.org/10.1117/12.319490

Eyre, J.R., English, S.J., Forsythe, M., 2020: Assimilation of satellite data in numerical weather prediction. Part I: The early years. Q J R Meteorol Soc. 2020; 146: 49– 68. https://doi.org/10.1002/qj.3654

Fois, F., 2015, Enhanced ocean scatterometry, PhD Delft University of Technology, Delft, the Netherlands, doi = 10.4233/uuid:06d7f7ad-36a9-49fa-b7ae-ab9dfc072f9c .

Kilic, L., Prigent, C., Boutin, J., Meissner, T., English, S., & Yueh, S., 2019: Comparisons of ocean radiative transfer models with SMAP and AMSR2 observations., J. Geophys. Res.: Oceans, 124, 7683– 7699. https://doi.org/10.1029/2019JC015493

As more than 70% of the earth surface is covered by water, exchanges of heat, gases and momentum at the air-sea interface are a key part of the dynamical earth system and its evolution. The ocean surface wind plays an essential role in the exchange at the atmosphere-ocean interface. It is therefore crucial to accurately represent the wind forcing in physical ocean model simulations. Scatterometers provide high-resolution ocean surface wind observations, but have limited spatial and temporal coverage. On the other hand, numerical weather prediction (NWP) model wind fields have better coverage in time and space, but do not resolve the small-scale variability in the air-sea fluxes. In addition, Belmonte and Stoffelen (2019) documented substantial systematic errors in global NWP fields on both small and large scales, using scatterometer observations as a reference.

Trindade et al. (2020) combined the strong points of scatterometer observations and atmospheric model wind fields into ERA*, a new ocean wind forcing product. ERA* uses temporally-averaged differences between geolocated scatterometer wind data and European Centre for Medium-range Weather Forecasts (ECMWF) reanalysis fields (ERA-Interim) to correct for persistent local NWP wind vector biases. Verified against independent observations, ERA* reduced the variance of differences by 20% with respect to the uncorrected NWP fields.

We present a new hourly ocean wind forcing product that will be included in the Copernicus Marine Service (CMEMS) catalogue. To best serve the ocean modelling community, this Level 4 product will include global bias-corrected 10-m stress-equivalent wind (De Kloe et al., 2017) and surface wind stress fields at 0.125 degree horizontal spatial resolution. The near real-time (NRT) version of the product is based on the ECMWF operational model (OPS*) and the reprocessed (REP) version on the ERA5 re-analysis (ERA5*). Like any CMEMS product, the new wind product will be freely and openly available for all operational, commercial and research applications.

References:
Belmonte Rivas, M. and A. Stoffelen (2019): Characterizing ERA-Interim and ERA5 surface wind biases using ASCAT, Ocean Sci., 15, 831–852, doi: 10.5194/os-15-831-2019.

Kloe, J. de, A. Stoffelen and A. Verhoef (2017), Improved use of scatterometer measurements by using stress-equivalent reference winds, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 10 (5), doi: 10.1109/JSTARS.2017.2685242.

Trindade, A., M. Portabella, A. Stoffelen, W. Lin and A. Verhoef (2020), ERAstar: A High-Resolution Ocean Forcing Product, IEEE Trans. Geosci. Remote Sens., 1-11, doi: 10.1109/TGRS.2019.2946019.

Local variability of sea surface wind has a significant impact on the mesoscale air-sea interactions and the wind-induced oceanic response, such as temperature variability and circulation patterns. Recent advances in the wind quality control of Advanced Scatterometer (ASCAT) show that wind variability within a wind vector cell can be characterized using certain quality indicators derived from ASCAT data, such as the inversion residual (namely the maximum likelihood estimator, MLE) and the singularity exponent (SE) derived from singularity analysis.

This study is aimed at quantifying the ASCAT subcell wind variability over the global ocean surface. It is assumed that the spatial variability is proportional to the variance associated with time-series of collocated moored buoys winds. As such, 10-min sampled buoy winds are used to examine the subcell wind variability following Taylor’s hypothesis, which allows for a temporal dimension to be converted into a spatial dimension, and vice versa. The time window (centered on the buoy measurement collocated with ASCAT acquisition) used for calculating the mean buoy winds and the subcell spatial variability is set equal to 25 km. Then the sensitivity of ASCAT quality indicators to the subcell wind variability is evaluated. The results indicate that SE is more sensitive than MLE in characterizing the wind variability, but they are rather complementary in flagging the most variable winds. Consequently, an empirical model is derived to relate the subcell wind variability to the ASCAT MLE and/or SE values.

Although the overall procedure is based on the one-dimensional temporal analysis and the empirical model cannot fully represent the two-dimensional spatial variability as depicted by the scatterometer, it is probably the first attempt to assign a subcell wind variability value for each wind vector cell within the ASCAT swath. The empirical method presented here is effective, straightforward, and could be applied to other scatterometer systems. The next step is therefore generate global wind variability maps which can be used in a wide variety of scientific and operational applications.

The presence of horizontal spatial structures in the sea surface temperature (SST) field is known to influence the atmospheric response at various time scales. The ESA CCI (Climate Change Initiative) GLAUCO (Global and Local Atmospheric response to the Underlying Coupled Ocean) project aims to characterize the wind, cloud and rainfall response to the SST structures at daily and sub-daily time scales, with a focus on the physical mechanisms responsible for that. In the literature, two main mechanisms have been identified: the Downward Momentum Mixing mechanism (DMM) mechanism and the Pressure Adjustment (PA) one. According to DMM, a positive change in SST along the wind decreases the stability of the lower atmosphere, which enhances the vertical mixing of horizontal momentum. This results in a net acceleration of the low level flow, producing wind divergence over SST fronts (from relatively cold water to relatively warm water). According to PA, the presence of a warm SST patch induces a local pressure low, responsible for pressure gradients that generate secondary circulations. Surface wind convergence (divergence) is then produced over local SST maxima (minima).
The long-term, consistent and unbiased climate data records produced within the ESA CCI project (and its extensions) are used. These data sets enable to robustly estimate the long-term statistics of the atmospheric response at fast (daily and sub-daily) scales. A “globally local” approach is pursued, where regional differences can be assessed and compared using observational products that are consistent at the global level. As different levels of processing are available, often associated with a different grid spacing, the dependence of the results on the size of the resolved SST structures can also be assessed. This can shed some light on the ability of general circulation models in representing these small-scale fast air-sea interactions and their impact on the atmospheric dynamics.
Typically, to determine the importance of the thermal mechanisms introduced above, one calculates correlation coefficients or the slope of the binned distributions (named coupling coefficients) of: downwind SST gradient and wind divergence for DMM, and SST Laplacian and wind divergence for PA. However, the advection has been observed to break the correlation between SST Laplacian and wind divergence, so that the PA mechanism has been often overlooked in the literature. As the pressure response is produced in all directions, we propose to measure the correspondence of the across-wind SST second spatial derivative and the across-wind divergence to identify the action of the PA mechanism. It is found that this new metrics detects a signal only when small-scale SST forcing is present.
By applying this new across-wind metrics to high resolution satellite data, namely the ESA CCI SST data at 0.05° and the L2 Metop-A ASCAT wind field swaths at 12.5 km, new interesting features of the wind response to the SST forcing appear. First of all, the signature of the PA mechanism appears in regions where this mechanism was thought not to be present. Moreover, both DMM and PA mechanisms show strong seasonal behaviors and regional differences. Non-linearities and asymmetries in the response according to the sign of the forcing field emerge, highlighting that the assumption of linear atmospheric response that has often been done does not hold at fine spatio-temporal scales. Thus, for a proper characterization of the air-sea fluxes, high-resolution simultaneous observations of SST, surface currents and MABL properties are needed, as pursued by the Earth Explorer 10 (EE10) candidate mission Harmony and the EE11 candidate mission Seastar.

Society requires decision-relevant, evidence-based climate information to support mitigation strategies and adaptation choices. It needs this to address the challenges created by the pace of climate change and the emerging risks associated with climate extremes and hazards, which are being exacerbated by climate change. Required for this is actionable climate information that is based on new knowledge resulting from improved observations, better process understanding, and im proved, robust predictions and scenarios of climate change, produced at increasingly fine spatial resolutions and over a wide range of timescales. While there remain key scientific gaps, there are also new opportunities to advance scientific understanding through strategic partnerships. The World Climate Research Programme (WCRP) stands ready to lead the way in climate science to advance climate knowledge in support of society by addressing frontier scientific topics related to the coupled climate system, supported by strong global partnerships and addressing scientifically and technically complex challenges to improve resilience and preparedness, mitigation and adaptation.

To meet the challenges and opportunities of climate science over the next decade, WCRP has developed a new Strategic Plan. WCRP’s new scientific objectives are informed by the most pressing contemporary climate knowledge needs, as well as advancing the core science and capability needed to prepare for the challenges that society cannot yet foresee. These objectives are: (1) to advance fundamental understanding of processes, variations and changes in the climate system; (2) to predict the near-term evolution of the climate system; (3) to refine the ability to anticipate future pathways of climate system change; and (4) to support the development of theory and practice in the integration between natural and social sciences. Through these objectives, WCRP’s Core Projects will contribute to progress in the foundations of climate physics and biogeochemistry, in the predictive skill across all climate system components, and in the improvement of simulations of the past and projections of the future. To rapidly advance the field for key scientific frontiers, WCRP has established Lighthouse Activities which are designed to be ambitious and transdisciplinary, integrating across WCRP and collaborating with partners. These Lighthouse Activities will also setup new institutional frameworks that are needed to manage climate risk and meet society’s urgent need for robust and actionable climate information more effectively.

The talk will summarize WCRP’s new strategy and approaches to advance and steer climate science in support of society and will highlight the role of observations and models in this context. To understand relevant processes and mechanisms in every component of the climate system, and therefore to fully understand the Earth system, we require a diversity of observations and modelling approaches that span a range of complexity, a range of representations of processes, and a range of spatial resolutions, to resolve progress down to small scales. Respective coupled climate processes are fundamental to understanding variations in large scale circulation, the trajectories of regional and global sea level rise, and extreme events, all of which have severe regional and local impacts. Closing the energy, water, and carbon budgets of these systems is integral to observing, assessing, and simulating climate change and variability, regionally and globally. Frameworks for model evaluation and uncertainty estimation are required, as is collaboration across model development communities. The potential for seamless and unified simulation tools, adaptive architectures, statistical methods, and machine-based learning is yet to be fully tapped, and new approaches are needed for the integration of observations and models towards better representation of the Earth climate system.

Climate change and environmental degradation are an existential threat to Europe and the
world. To overcome these challenges, the European Green Deal will transform the EU into a
modern, resource-efficient and competitive economy.

From November 2020 to July 2021, Eurisy and DotSPACE hosted a webinar series bringing
together research, government and industry experts to talk about their innovative solutions
related to climate. Throughout the Space Opportunities for Climate Challenges series, various
examples have been showcased proving how satellite solutions can empower the green
transition. For example, satellite remote sensing can rapidly reveal where to reverse the loss
of biological diversity. Variables such as vegetation productivity or leaf cover can be measured
across continents from space and can help forest managers to implement more sustainable
ways of working. Furthermore, space is relevant for the management of maritime-related
matters, as it is for smart mobility and urban planning. When it comes to energy, space can
play a pivotal role in the decarbonisation of our economy. The daily space data stream also
provides insights about air and water quality, as well as irrigation systems, and even tourism.

Despite their strategic importance for sustainable economic, environmental and social
development, the uptake of satellite-based services remains limited. A lack of awareness of
the availability of satellite data and its potential interoperability with local data still poses a
major obstacle. Mitigating this issue will be crucial since impacts of climate change are
increasingly being felt at local level. Eurisy’s database of success stories is one of the
association’s primary tools to promote the use of satellite solutions for the benefit of
professional communities in numerous sectors on national, regional and local levels. By
sharing hands-on experience of public authorities, agencies, and SMEs, Eurisy aims to
eventually reach a critical mass of users and decision-makers tapping into the space data
stream to implement climate adaptation policies more easily.

Reanalyses form a key component of the suite of data products developed by the Copernicus Climate Change Service (C3S). ECMWF’s fifth generation global reanalysis (ERA5) has supported a growing user base, currently numbering more than 50 000 users. A preliminary extension of ERA5 back to 1950 was published in 2021. In addition to assimilating a comprehensive set of satellite data starting from the TOVS suite of instruments from 1979 onwards, ERA5 makes use of early infrared sounding data from VTPR, carried on NOAA-2 through-5 from 1972-1979. Reprocessed satellite data have played a role, alongside model and data assimilation developments, in delivering improved analyses relative to ERA5’s predecessor, ERA-Interim. Preparations are now underway for the next generation of reanalysis, ERA6, due to start in early 2024.
ERA6 will make use of several reprocessed satellite datasets produced by EUMETSAT as part of the first (2015-2021) and second (2021-2028) phases of the EU’s C3S programme. Plans currently include the production and assimilation of Fundamental Climate Data Records (FCDRs) for ATMS, MHS, MWHS-2, HIRS, SSM/T, SSMIS and European (MVIRI and SEVIRI) and Japanese geostationary satellite radiances. The first phase of the C3S programme has also delivered reprocessed datasets for several radio occultation missions (GRAS, COSMIC, GRACE and CHAMP) as well as Atmospheric Motion Vectors (AMVs) and scatterometer data (ASCAT). This element of C3S aims to produce comprehensive uncertainty analyses for the microwave sounders MSU, AMSU-A and ATMS. The impact of these new reprocessed datasets has been assessed in observing system experiments (OSEs) at ECMWF which show the new data generally exhibits lower biases, results in improved re-forecast quality and, in some cases, have an impact on the mean state estimate.
ERA6 will also make use of several recently rescued early (1970s) satellite datasets, including radiances from SI-1, SMMR, SSH, IRIS, SIRS, PMR, MRIR, NEMS, SCAMS, ESMR and SCR. Preparations to date have included the generation of improved radiative transfer models and evaluation of the quality of these radiances relative to ERA5 using analysis departures computed off-line.

Among the several effects of climate change, the rise of oceans level and occurrence of extreme meteorological events will inevitably result in coastal flooding episodes, temporarily or permanently eating away the coastline.
Rising to the challenge of mapping coastal flooding hazards and assessing their socioeconomic risks from satellite, the Littoscope project, supported by the French Space Agency in the framework of the international Space Climate Observatory (SCO), promotes the use of satellite data for information and decision-making related to the impact of rising oceans in coastal areas.
Littoscope is based on three pillars: a) mapping the coastal flood hazards using high-resolution optical satellite images and satellite altimetry data to estimate future flooded areas, b) assessing coastal flood risks based on local exposures and c) establishing an IT tool dedicated to local decision-makers.
The mapping of several scenarios of coastal flood hazards relies on Pleaides satellite images, from which is derived a high-resolution (0.5m) Digital Elevation Model (DEM). It takes into account the sea level trends estimated from satellite altimeter missions since 1993 (data from the Copernicus Climate Service) and decadal intensity of storm and tide surge from the global model MOG2D. The potential flooding water heights are first estimated through a bathtub approach comparing oceanic and terrestrial heights. A high-resolution hydrodynamic model is also applied to evaluate the capability of satellite-derived DEM to be used in climate modeling study or early warning tools to prevent coastal flooding risks.
The resulting risks of coastal flooding are estimated by crossing the hazard intensity with social, economic, natural and cultural exposures coming from a multi-sourcing (national, regional and local GIS datasets combined with land use information derived from the HR optical satellite image).
Two demonstration areas have been selected in France: the peninsula of Gâvres (4.5 km²) in Brittany and a broader area (102.4 km²) around the ponds near Palavas-les-Flots on the Mediterranean coast. The web platform has been co-designed with its future users among these territories to provide them with a comprehensive and easy-to-use tool.
This paper will exhibit the main outcomes of the Littoscope project and demonstrate how satellite assets can be used in combination with ocean models and socio-economics data to propose an highlighting tool about climate risks in the coastal areas.

Snow cover and Lake Ice cover have been both specified by the Global Observing System for Climate (GCOS) as part of the 50 essential climate variables (ECVs) to be monitored by satellite remote sensing. They are relevant input parameters for forecasts in the field of weather, hydrology and water management, therefore essential for assessing natural hazards such as floods, avalanches or river ice jams and managing associated risks.
Since July 2020, under European Environment Agency (EEA) delegation, the Copernicus Land Monitoring Service (CLMS) operationally produces and disseminates Pan-European High-Resolution Snow & Ice products (HR-S&I) at high spatial resolution (20 m x 20 m and 60 m x 60 m). They are derived from high-resolution optical and radar satellite data, from the Sentinel-2 and Sentinel-1 constellations respectively. Among near-real time (NRT) products, snow properties are described by two types of them.
Snow cover:
The Fractional Snow Cover (FSC) product provides the snow fraction at the Top Of Canopy (FSCTOC) and On Ground (FSCOG).
The daily cumulative Gap-filled Fractional Snow Cover (GFSC*) product provides a more complete FSC product, gap-filled both at spatial and temporal scales.
Snow state:
The Wet/Dry Snow (WDS) product differentiates the snow state within the snow mask defined by the FSCTOC information.
The SAR Wet Snow (SWS) product provides information on the wet snow extent in high-mountain areas.
Ice occurrences on the European hydrographic network are described by the River and Lake Ice Extent (RLIE) product. There are several RLIE products available, depending on their data source (either Sentinel-1, Sentinel-2 or a combination of both types of observations).

HR-S&I products are generated over the entire EEA38 (32 member countries and 6 cooperating countries) and the United Kingdom. The Sentinel-1 archive data from September 2016 is currently being processed while Sentinel-2 based products are already available from September 1, 2016 onwards to users.
This new Copernicus Service component has been developed and is currently operated by a consortium led by Magellium in partnership with Astri Polska, Cesbio, ENVEO, FMI and Météo-France, contracted by EEA as entrusted entity of DG-DEFIS (European Commission - Directorate General Defense Industry and Space) for this service.
The service is based on preexisting Research & Development algorithms and products conducted by CESBIO and supported by CNES for FSC products (MAJA(1) and LIS(2)), conducted by ENVEO and FMI teams for wet snow and GFSC products(3,4), and by Astri Polska developments for ice products. These algorithms have been turned into operational conditions, based on the WEkEO DIAS European cloud infrastructure.
Maximum efforts are made to provide NRT HR-S&I products within the ideal user requirement timeliness, ie 12 hours after the sensing date the maximum acceptable time lag for time-critical applications such as avalanche bulletins, weather forecasting, etc. This timeliness depends mainly on the date of publication of Sentinel-2/1 data on the Copernicus Service Hub by ESA, then once this data is available, NRT HR-S&I products are published within a maximum of 3 hours.

The presentation aims at describing in detail this new component of the CLMS.


(1) MAJA: the MACCS-ATCOR Joint Algorithm provides atmospheric correction and cloud-screening module to generate L2A products. It is a software developed by CNES and CESBIO, with contributions from DLR (https://zenodo.org/record/1209633#.XoGq7vE69hE)
(2) LIS: “Let It Snow” Snow detection algorithm: Gascoin, S., Grizonnet, M., Bouchet, M., Salgues, G., and Hagolle, O. (2018) Theia Snow collection: high resolution operational snow cover maps from Sentinel-2 and Landsat-8 data, Earth Syst. Sci. Data, https://doi.org/10.5194/essd-11-493-2019
(3) Nagler T. and H. Rott. 2000. Retrieval of wet snow by means of multitemporal SAR data. . IEEE Transactions on Geoscience Remote Sensing. vol. 38, no. 2, pp. 754–765, Mar. 2000.
(4) Nagler T., H. Rott, E. Ripper, G. Bippus, and M. Hetzenecker. 2016. Advancements for Snowmelt Monitoring by Means of Sentinel-1 SAR. Remote Sensing, 2016, 8(4), 348, DOI:10.3390/rs8040348.

The Copernicus Land Monitoring Service (CLMS) is one of the six core user driven services of Copernicus, the European flagship programme on Earth Observation and focuses on land monitoring, specifically at European scale, and also at global scale. The Land Service operationally produces a series of qualified bio-geophysical products on the condition and evolution of the land surface and on land cover and land use status and change. EU scale products are at high spatial resolution and global scale products are at mid to low spatial resolution with high time frequency; all are complemented by the constitution of long-term time series. The products are used to monitor the changes in dynamics of vegetation, the variability of the water cycle, the land energy budget, the terrestrial cryosphere and land cover changes. The service supports an increasing global user community for land resource monitoring, planning and also supports activities related to climate change monitoring and mitigation and adaptations initiatives. COP26 highlighted that, although the terrestrial biosphere feedbacks are a main issue, the impacting role of land cover changes and land management interventions still has a low level of confidence. GHG budgets over land are still ‘weak’ elements in climate models, hence detailed, correct and dynamic land data and information is more urgently required for climate related activities.
Space borne earth observations, however, directly support the development of bottom-up emission inventories for agriculture, forestry and other land uses (AFOLU). Such data also offer the capacity to monitor changes of land cover, land use and their biophysical condition that allow identification of options for creating resilience to destructive climate change impacts.
The need for continuous and quality controlled land data is recognized by IPCCC. The Copernicus Land Service global component therefor collaborates with the CEOS initiative in support of the UNFCCC Global Stocktake, scheduled for 2023.
We present details and applications of the CLMS products that are contributing to this and highlight compatibilities and differences with products from the Climate Service. We illustrate the value of integrated and derived datasets, such as water quality, land productivity and land degradation pressures. Furthermore, CLMS introduced processes that focus on user involvement, an important aspect in the MRV process (Measuring, reporting, verification), such as products linked to REDD+ for forest monitoring and GEOGLAM for agriculture and BIOPAMA for biodiversity.

1. INTRODUCTION

The European Plate Observing System (EPOS) ‎[1] is a long-term plan to foster and facilitate the integrated use of data, products, software and services made available through distributed European Research Infrastructures (RI) in the field of Solid Earth Science (SES). In particular, EPOS is a pan-European Research Infrastructure of the ESFRI Roadmap and it has been recently established with an ERIC (European Research Infrastructure Consortium) hosted by INGV in Italy. EPOS is supported by 25 European countries and several international organizations.
EPOS integrates a large number of existing European RIs belonging to several fields of SES, referred to as Thematic Core Services (TCS), from seismology to geodesy, near fault and volcanic observatories, as well as anthropogenic hazards and satellite observations. The EPOS vision is that the integration of the existing national and trans-national RIs will facilitate the access and use of the multidisciplinary data recorded by the solid Earth monitoring networks, acquired in laboratory experiments and/or produced by computational simulations. The EPOS establishment will foster the interoperability of products and services in the Earth science field to a worldwide community of users. Accordingly, EPOS aims at integrating the diverse and advanced European RIs in the field of Solid Earth Science and building on new e-science opportunities to monitor and understand the dynamic and complex Solid Earth System. One of the EPOS TCS, referred to as Satellite Data (SATD), aims at developing, implementing and deploying advanced satellite data products and services, mainly based on Copernicus data (namely Sentinel acquisitions), suitable to be largely used by the SES community.
For a research infrastructure supporting a large community from multiple countries, as the case of EPOS, it is critical that the underlying infrastructure and the computing resources carefully take into account the operational environment for users and its sustainability, both technical and financial. This point is particularly relevant for the SATD RIs that, to deploy robust and effective services, have to properly manage several issues related, for example, to the satellite data access, archive handling and storage, the management of the computing facilities, and the efficient and automatic processing chains. In this framework, the European scenario is rapidly evolving and several pan-European initiatives have been recently fostered. Among all, the European Open Science Cloud (EOSC) ‎[2] and the Copernicus Data and Information Access Services (DIAS) ‎[3] platforms represent the most promising opportunities to reach the long-term technical sustainability of the EPOS TCS SATD.
This work is focused on the technological enhancements and the activities carried out to implement the TCS SATD and to deploy effective EO satellite services in a harmonized and integrated way by benefitting from the current and future European satellite scenario. In particular, we present the advanced Differential SAR Interferometry (DInSAR) techniques implemented to provided robust services to map and investigate the ground motion of local and wide scale deformation phenomena, from regional to continental analysis. Finally, we show the procedures and methods developed or adopted by the TCS SATD to guarantee good data management and stewardship by following the FAIR principles.

2. EPOS TCS SATELLITE DATA
The structure of the EPOS TCS Satellite Data is shown in Figure 1. The scope of this TCS is the implementation of Earth Observation services, based on satellite observations, transverse to the large EPOS community and suitable to be used in several application scenarios. In particular, the main goal is to contribute, with mature services that have already well demonstrated their effectiveness and relevance in investigating the physical processes that control earthquakes, volcanic eruptions and unrest episodes, as well as those driving tectonics and Earth surface dynamics. The development of TCS SATD is supported by 5 European institutions providing different services (Table I), CNR and INGV (Italy), CNRS (France), CSIC (Spain) and University of Leeds (UK), and benefits from the collaboration with the European Space Agency (ESA).
At this stage, two levels of products and services, based on Differential SAR Interferometry (DInSAR) techniques ‎[4] to estimate and analyze Earth surface displacements and terrain motion mapping for geohazards applications, are distributed. The first level deals with “standard” satellite products/tools (e.g., SAR interferograms, LOS displacements maps and deformation time-series generation). The second level concerns value-added satellite products/tools (e.g., modelling analyses, 3D displacement maps, source mechanisms, fault models, strain maps). The TCS services are mainly based on Copernicus data (Sentinel-1/2 datasets); in addition, advanced DInSAR web processing services dealing with ERS-1/2, ASAR-ENVISAT, and Sentinel-1 data are made available by the TCS SATD. Since the services include both access to products and processing utilities, we have to consider two specific functioning modes:
• Continuous mode - systematic and periodic generation of products (e.g. the systematic production of updated surface deformation time series over given areas);
• On-demand mode - users run the tools and process chosen satellite datasets (e.g. ad hoc generation of deformation measurements using satellite observations during a telluric crisis, such as a co-seismic motion map).
The continuous services systematically generate products directly accessible by the users. Such products are relevant to areas of the Earth surface significant for the Solid Earth Science community (volcanoes, faults, seismogenetic areas, geohazard supersites, etc), while the on-demand services make available to users advanced web-tools able to generate satellite products by processing Copernicus datasets (Figure 2).
The TCS community worked to apply FAIR principles [6] to its products and data. Following OGS guidelines, the TCS implements the ISO 19115 standard, that has been adapted to describe interferometric SAR products, both maps and time series. Moreover, all the TCS products are distributed following the open data and open access policy, with CC-BY license.
The TCS has a unique thematic interface towards the EPOS central hub, referred to as Integrated Core Services (ICS) (Figure 1). This interface is represented by the Geohazards Exploitation Platform (GEP) ‎[5], developed with the support of ESA, which is able to provide interoperable access to data products, web processing tools and processing facilities. This unique gateway provides the user interface (GUI) and the interoperability layer of the TCS, establishing unique AAAI and API for the several RIs.

REFERENCES
[1] European Plate Observing System (EPOS) [Online] Available: http://www.epos.org
[2] European Open Science Cloud (EOSC) [Online] Available: https://www.eosc-hub.eu
[3] Copernicus Data and Information Access Services (DIAS). [Online] Available: http://copernicus.eu/news/upcoming-copernicus-data-and-information-access-services-dias
[4] A. K. Gabriel, R. M. Goldstein, and H. A. Zebker, “Mapping small elevation changes over large areas: Differential interferometry,” J. Geophys. Res., vol. 94, no. B7, pp. 9183– 9191, Mar. 1989.
[5] Geohazards Exploitation Platform (GEP) [Online] Available: https://geohazards-tep.eu/
[6] Wilkinson, M., Dumontier, M., Aalbersberg, I. et al., “The FAIR Guiding Principles for scientific data management and stewardship,” Sci Data, vol. 3, n. 160018, 2016. https://doi.org/10.1038/sdata.2016.18

Differential Synthetic Aperture Radar Interferometry (DInSAR) has extensively proven in the last decades its unprecedented capability to measure Earth surface displacements at very-large scale and with high accuracy. In particular, DInSAR techniques allow us to retrieve ground deformation related to both natural hazards (earthquakes, tectonic movements, volcanic phenomena, landslides, etc.) and anthropic actions (mining, gas injection/extraction, groundwater exploitation, etc.) with centimeter to millimeter accuracy [1]. Moreover, the current DInSAR scenario is characterized by a huge availability of SAR data acquired by the large number of operating SAR sensors, such as ALOS-2, COSMO-SkyMed, RADARSAT-2, SAOCOM, Sentinel-1, TerraSAR-X, which will be further increased thanks to the already planned NISAR and ROSE-L missions.
Actually, one of the main limitations for correctly retrieving deformation signals from DInSAR results is the Atmospheric Phase Screen (APS) component, which accounts for the presence of atmospheric artifacts contaminating the intereferometric measurements. Indeed, since atmospheric properties such as temperature, pressure and humidity can vary in space and time, the refractivity index of the atmosphere (through which the transmitted microwave pulses and the backscattered signals propagate) may change between the acquisition times of the two SAR images resulting in an interferogram. Consequently, the generated interferogram will present the above mentioned APS component not related to deformation. Moreover, there can be scenarios in which distinguishing the APS from the real deformation is particularly complicated. This is, for instance, the case of areas characterized by the presence of significant topography (since atmospheric properties vary with height producing a topography-correlated atmospheric phase component) and significant deformation as well, as in the case of volcano eruptions. This is particularly true when the displacement component is comparable to the one owing to the atmospheric artifacts [2].
A significant number of methods for APS mitigation have been proposed in literature over the last years, which can be essentially classified in two categories: DInSAR time series and external data based approaches. The first solutions account for the APS statistical properties both in space and in time in order to filter out atmospheric contributions from DInSAR time series, and is typically effective with large datasets. The second techniques rely on the use of external auxiliary data (e.g. Zenith Total Delay from GPS measurements, meteorological models or Global Atmospheric Models (GAM)) in order to directly estimate and remove the atmospheric artifacts from interferograms; the major drawbacks, in this case, are represented by the generally low spatial and/or temporal resolutions of the external data. However, thanks to the development of numerical weather prediction models over recent years, we have now available GAM datasets providing accurate atmospheric parameters measurements having rather high resolutions. In particular, the ECMWF ERA-5 datasets, generated by exploiting the Copernicus Climate Change Service Information [3], are available on a global-scale, covering the Earth on a 30km grid. In particular, the ERA-5 data resolve the atmosphere using 137 levels from the surface up to a height of 80km and are available hourly.

In this paper we take in considerations three sites particularly challenging from the point of view of the APS estimation and removal, because of the complexity of the investigated scenarios:
i) the Canary island of La Palma (Spain), located in the Atlantic Ocean, which is a volcanic complex characterized by a long eruptive activity, focusing on the last eruption occurred on 19 September 2021 that is still ongoing and to date has caused extensive damage to homes, infrastructures and many productive activities on the island;
ii) the volcanic area located in the Napoli bay area (Italy) that includes the Vesuvio Volcano and the caldera of Campi Flegrei, the latter being characterized by the bradyseism phenomenon that caused in the last years an uplift with a rate reaching 10 cm/year;
iii) the Etna Volcano (Sicily), which is the Europe's largest and most active volcano, characterized by frequent eruptions often accompanied by large lava flows.

The three above-mentioned sites are all characterized by relevant deformations and significant heights, therefore it is often difficult to separate the APS topography-related interferometric component from the actual ground displacement causing an underestimation of the last one.
The aim of the work is to evaluate and compare the performances of the APS filtering solutions:
1) exploiting the join availability of spatial and temporal information given by long deformation time series generated through Sentinel-1 data acquired over the sites of interest,
2) applying to the DInSAR interferograms stacks the APS corrections directly calculated by exploiting the external ERA-5 data [4].
In the final presentation we will show a comprehensive analysis of the results obtained through the above mentioned two different approaches, and achieved both on individual interferograms and on deformation time series. Moreover, we will also investigate the possibility to combine the two techniques, in order to estimate and remove the APS interferometric component in a more accurate way. Finally, the impact of using the experimental Sentinel-1 ETAD products, which account for tropospheric and ionospheric corrections [7], will be possibly analyzed.



References

[1] D. Massonnet and K. L. Feigl, “Radar interferometry and its application to changes in the Earth’s surface,” Rev. Geophys., vol. 36, no. 441-500, 1998
[2] Zebker, H. A.,Rosen, P. A., and Hensley, S. (1997), Atmospheric effects in interferometric synthetic aperture radar surface deformation and topographic maps, J. Geophys. Res., 102( B4), 7547– 7563, doi:10.1029/96JB03804
[3] https://cds.climate.copernicus.eu/
[4] Jolivet, R., P. Agram and C. Liu, Python-based Atmospheric Phase Screen estimation - User Guide (2012), http://earthdef.caltech.edu .
[5] Casu, F., Elefante, S., Imperatore, P., Zinno, I., Manunta, M., De Luca, C. , and Lanari, R., SBAS-DInSAR Parallel Processing for Deformation Time-Series Computation. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., vol. 7, no. 8, pp. 3285–3296, 2014
[6] Manunta, M.; De Luca, C.; Zinno, I.; Casu, F.; Manzo, M.; Bonano, M.; Fusco, A.; Pepe, A.; Onorato, G.; Berardino, P.; De Martino, P.; Lanari, R., The Parallel SBAS Approach for Sentinel-1 Interferometric Wide Swath Deformation Time-Series Generation: Algorithm Description and Products Quality Assessment, IEEE Trans. Geosci. Remote Sens., 2019, 57
[7] https://sentinels.copernicus.eu/web/sentinel/missions/sentinel-1/data-products/etad-dataset

The “Pas de l’Ours” landslide, located in the Queyras valley (Southeast France) is undergoing periods of fast deformation that began in the Spring 2017. The total moving mass is estimated at 17 million cubic meters, with a width of 1 km and a length of 600 m, which makes it currently one of the largest active landslides in the French Alps. In addition to the large deforming mass, numerous rockfalls and mudflows have occurred and have severely damaged the road that is located at the foot of the landslide.

Ground-based instruments including a GBSAR, GNSS, and seismometers have been deployed on-site to monitor the landslide evolution. In addition, we are monitoring the landslide motion using SAR data from the Sentinel-1(S1) satellite and optical images from Sentinel-2 (S2) and Planet. These various datasets provide us with multiple measurements that can be combined to better understand the complete landslide behavior.

We analyze the landslide deformation pattern measured by the satellite images and ground-based InSAR techniques between 2015 and 2019. The SAR acquisitions are processed using the SqueeSAR algorithm and ISCE. The optical images are processed with the MPIC-OPT-SLIDE service available on the ESA Geohazards Exploitation Platform (GEP). These measured derived from satellite acquisitions are compared with the in-situ measurements (GB-InSAR and GNSS’s). We show the complementarity of these techniques to measure different kinematic regimes from tens of millimeters per year to tens of meters per year. We are able to detect the onset of the acceleration of 2016-2017 and discuss the triggering factors. We also investigate the progressive decrease of the magnitude of the following accelerations (2018, 2019). The location of the active parts of the landslide also varies through time. This work highlights how complementary monitoring techniques can be combined to retrieve the evolution of the slope instabilities in various kinematic regimes, and offers a new perspective into the physical processes controlling landslide dynamics.

Accurate and timely observation of mountain processes is considered one of the most important scientific activities in the field of environmental sciences in recent years, as it allows to detect indications and/or precursors of changes potentially impacting ecosystems at regional and global scales. One of the peculiarities of geologically young mountain ranges is the possibility to directly observe and study the evolution of paraglacial and periglacial morphological processes, i.e. all earth surface modifications that are directly conditioned by cyclic glaciation and deglaciation periods. Because of the current climatic changes, modifications of the mountain landscape due to large mass movements (i.e., glaciers, landslides of different size and typology and rock glaciers) are more and more observed and their impacts are expected to increase. In some cases, the rapid and potentially catastrophic evolution of such mass movements might directly affect anthropogenic infrastructure, economic activities, and also human lives. Hence, systematic mapping and monitoring of existing slope instabilities as well as the investigation of past catastrophic slope failures are essential for an effective hazard assessment, risk management and disaster response.

In this contribution, we show the initial results obtained by processing and analyzing available satellite radar datasets acquired from the ESA Sentinel-1 mission in the period 2018-2021. We focused on the Bhagirathi Valley, Uttarakhand, India, a high alpine area more and more threatened by large and catastrophic slope collapses. Currently different hydropower projects are under construction or in the planning phase within the region, increasing infrastructure at risk and the potential for cascading disasters such as the Chamoli rock/ice avalanche in 2021. We use standard radar interferometry to first detect and classify areas affected by potential instability. In addition, we focus on specific locations to determine the spatial and temporal evolution of surface displacements and to identify potential changes of trends associated to climatic variables. The result of this large scale and systematic investigation will be the base to test and calibrate numerical models of mass movements. With these models, the simulation of rock-, ice- and snow avalanches, as well as processes combining these input materials, enables the assessments of the hazard intensities and the generation of hazard indication maps for the region. These are essential tools for the planning of effective mitigation measures such as hazard zonation.

The last years, impacts of natural disasters on populations and infrastructures are rising worldwide. All efficient solutions are therefore required to support hazard and risk assessments, land-use planning, public risk financing and disaster forecasting through integrated landslide risk management systems. For landslide disasters, effective and extensive products are challenging at regional and country scales, especially when robust landslide inventories, appropriate destabilization factors and triggers or exposed populations and infrastructures must be captured and post-processed.
In the same time, the acquisition, quality and access to satellite optical and radar observation data (with for example Copernicus, Pléiades or TerraSAR missions) and rainfall measurements (such as Global Precipitation Measurement or Global Forecast System missions) significantly progressed the last decade. In addition, robust and efficient models are now trained enough to support near-real time landslide and risk assessments, with for example the Landslide Hazard Assessment for Situational Awareness (LHASA), the FLOW path assessment of gravitational hazards at a Regional scale (FLOW-R) or the Potential Impact Index (PDI) models.
This presentation aims then to present the prototype App LHIS, for LANDSLIDE HAZARD INFORMATION SYSTEM. Indeed, LHIS promotes the landslide awareness and disaster risk financing by providing an App to anticipate, forecast and respond to incipient landslide events in near-real time. LHIS is based on integrated landslide-related models (ALADIM, LHASA, FLOW-R, PDI) automatically executed online and applied on HR satellite imagery. Its implementation as a web-service within the Geohazard Exploitation Platform (GEP) allows an easy access, processing and visualization of these EO-derived products. LHIS targets two main usage modes: LHIS-Nowcast and LHIS-Impact.
LHIS-Nowcast aims to forecast in near-real time the landslide hazard triggered by extreme rainfall events and identify and quantify related exposed infrastructures and populations. Based on the chained LHASA, Flow-R and PDI models, its indeed models the failure susceptibilities of potential landslide sources identified according to rainfall nowcasts and computes their maximum propagations. Then it identifies related exposed populations and infrastructures that can be reached and estimates the potential damage costs.
LHIS-Impact aims to map and assess damages in the aftermath of a major landslide event. Based on the chained ALADIM and PDI models, it can detect changes on HR to VHR optical satellite images before/after a specific selected event and automatic delimitation of the impacted area. Then it identifies the affected infrastructures and populations and computes the real damage costs.
Together, both LHIS usages modes can therefore support land-use mapping, for cost/benefit analyses of prevention measures), as well as estimation of the insurance coverage needed for reconstruction, by providing pertinent results for parametric insurance calculations, including landslide inventories, susceptibility and hazard maps, potential damages and costs analyses in near-real time, and real damages and cost after a major landslide disaster.
The prototype was developed and tested in Morocco in close collaboration with the FSEC (solidarity fund against catastrophic events) and the World Bank. The input data, processing details and results, calibrated for the pilot study on the Rif Tangier-Tetouan peninsula (Northern Morocco), will be presented.

Following a seismic swarm that started on 11 September 2021 and gradually intensified, a magma pathway propagated along the Cumbre Vieja rift zone of La Palma. On 19 September 2021, the eruption began by the opening of an 800 m long fissure located in the area of Cabeza de Vaca, El Paso, on the mid-western flank of Cumbre Vieja. The eruption intensified over the next weeks and was characterized by lava fountaining from multiple vents, Strombolian explosions, and advancing lava flows towards the western coast of the island. Thereby, over 2,600 houses were destroyed. On 28 September, the lava flows entered into the ocean and initiated the formation of a lava delta that is episodically growing. Due to the high impact hazards, the area is hardly approachable by field sensors allowing estimation of the dimension and evolution of surface changes. Therefore, we have been acquiring and analyzing multiple sensor satellite data. During the initial dike intrusion period, multi-temporal differential SAR interferometry analysis of C-band (Sentinel-1) and X-band SAR sensor data (PAZ, TerraSAR-X/TanDEM-X, Cosmo-SkyMed) showed over 40 cm deformation of the affected slope towards the sea. The following eruption was monitored by SAR amplitude data, VHR and HR optical satellite data (Pléiades, GeoEye, Sentinel-2, Landsat-8, hyperspectral DESIS) to study the spatio-temporal evolution of lava flows and ash deposition. At the time of writing (15 of November 2021), the lava flows of the still ongoing eruption covered an area of over 10 km². The growing lava delta accumulated to an area of ~0.4 km². Thermal data of Moderate Resolution Imaging Spectroradiometer (MODIS) and Visible Infrared Imaging Radiometer Suite (VIIRS) have been jointly analyzed to estimate the temporal evolution of the lava effusion rate and the total volume (6.22 × 10^7 ± 3.11 × 10^7 m³ by mid of November 2021). Although this thermally identified volume is already close to those of other historical eruptions on La Palma, it is largely underestimating the real volumes that erupted in 2021. Monitoring by the mid-resolution thermal sensors MODIS and VIIRS has the advantage of a high observation frequency, but only lava emplaced at the surface during the time of the satellite overpass can be detected. The difference of the thermally identified lava volumes to estimates arising from field observations may provide a hint of the importance of hidden lava flows. Lava flowing underground and in tubes and directly entering the ocean cannot be detected by these sensors. Therefore, change analysis between pre-eruption digital elevation models (DEMs) and newly created co-eruption DEMs from bi-static TanDEM-X data as well as from Pléiades stereo data are essential to understand the subaerial and hidden lava flow dynamics and to derive the total erupted lava volume.

Copernicus4regions - How interregional best-practice and knowledge sharing contributes to space capacity building
Roya Ayazi1, Margarita Chrysaki1, Branka Cuca2
NEREUS – Network of Regions Using Space technologie1, Dept. of Architecture, Built environment and Construction engineering (DABC), Poltecnico di Milano2

The Copernicus4regions campaign comprises 99-user-stories on how SENTINEL-imagery is successfully used at local and regional level. It is a unique ongoing interregional cooperation at European scale towards the common goal of bringing the benefits of the system to the regional and local level while expanding to new user-communities. It is a joint initiative of the European Commission, the European Space Agency and the European Network NEREUS. Community building and sharing experiences, knowledge and best practices to make more and better use of Copernicus are core elements of the initiative. It is also a step to bridge regional users with the political level and raise awareness amongst European politicians for the societal value of the system.

It is a truly bottom-up approach that builds on story-telling and was realized by volunteers from different disciplines and sectors across Europe. The broad range of authoring backgrounds and organisations who contributed to the collection shows how SENTINEL imagery is now diffusing into society at all levels. The use-cases that are subject of the collection were received in response of a call in 2017, open to all Copernicus Contributing Countries. It invited contributions in 8 selected application domains with high relevance and closely linked to the competences of the local and regional level. In total, the stories cover 177 authoring entities, 72 regions of applications and 24 European countries. The affiliation of authors who come from 28 different European countries reflect a rich geographical and structural diversity of the expertise in developing and handling Copernicus-based solution. This might also hint to an increased spread of skills and capabilities across Europe.

In fact, local and regional authorities, who contributed considerably and were quoted in the majority of use-cases, are the focus of the campaign. The overall idea had been to make the deployment situation at the level of local and regional administrations more transparent and empower public authorities to learn from each other and partner up. Public authorities share manifold challenges for which the program provides solutions and new approaches. Despite the fact that there is a positive trend in the up-take, public authorities, the main users and customers of Copernicus services, still have to overcome a variety of obstacles to fully exploit the benefits and potentials of the Copernicus-ecosystem.

By analysing how and to what extent other regions have tackled common challenges, Copernicus4regions identified a number of positive use-cases that exemplify the benefits of the Programme are suited to serve as a positive model to other regions. These clear and in-depth portrayals of users’ stories are meant to motivate regional stakeholder to explore use opportunities and get involved.

In this vein the collection is meant to showcase on the one hand the process of transforming data into valuable information for public authorities with tangible benefits for regions and their citizens but also on the other hand to highlight innovative processes and sustainable mechanisms that lead a public administration to develop and use a space-based product and/or service. In this respect Copernicus4regions contributes to space capacity building in regions in many respects.

With local and regional authorities being the protagonists of the collection, public sector innovation is a key topic: To this end the practical reference to user experiences in public policy and territorial contexts demonstrates how the data can be used to modernise and innovate the public sector while providing more efficient public services, improving the quality of life and level of satisfaction for European citizens.

The vast majority of the papers received are cases describing mature fields of application for satellite Earth observations such as “Agriculture, Food, Forestry and Fisheries” (32) followed by Biodiversity and Environmental Protection (17). Most of the user stories refer to data from Sentinel-1 (mentioned in 44 user stories) and Sentinel-2 (74).

Bearing in mind, that strong political will and commitment of civil servants is an important factor to pave the ground for an integration of Copernicus’ services into the workflows of public administrations, the campaign specifically targets policy makers and public authorities as mentioned above. For this purpose, targeted outreach/promotional tools were developed to make the collection more attractive and comprehensive to these specific groups. Besides Copernicus4regions offered a forum to regional users and politicians to exchange and debate regional use cases and deployment situation. The campaign organized a number of events to bring tangible user-experiences from the local and regional level to the European Parliament and sensitize politicians and their staff for the program and its impact on society. These meetings were also important occasions for regional user representative to liaise with other relevant stakeholder and voice their views and needs towards the political and institutional communities. Given the restrictions by the pandemic the Copernicus4regions community continued its efforts and dialogue via targeted webinars that put the focus on bringing new stories to the stage around the Green Deal Agenda of the European Commission.

The publication “The Ever Growing Uses of Copernicus across Europe’s Regions” and additional outreach material that complements the collection are freely downloadable from the NEREUS website. The Copernicus4regions outreach material comprises a brochure, single info-sheets, search-engine with different parameters, teaser, videos and webinars.

The Copernicus4regions collection builds on former experiences such as the 2012 publication “The Growing Use of GMES across Europe’s Regions” (67 user cases) and SENTINEL4regions, “Improving Copernicus take-up among Local and Regional Authorities via dedicated thematic workshops” (2015/16). In order to capture the evolutions of the 99 stories since the release of the publication in 2018, the organizing team launched a consultation of authors in 2021 to gain information in how far the use-cases where integrated into workflows of public administration, if the solution was institutionalized and analyse its evolution, to assess any technological improvements, but most importantly to identify possible additional benefits to the public administrations sector and citizens, that have been observed and determined over the past few years (on-going activity). The idea is to analyse the evolution of the whole collection of Copernicus4Regions User Stories in a systematic way, monitoring the changes and novelties.

Acknowledgements: This activity was managed by the Network of European Regions Using Space Technologies (NEREUS) under a contract from the European Space Agency. The activity is funded by the European Union, in collaboration with NEREUS. Paging, printing and distribution of this publication is funded by the European Space Agency.

As noted in Sathyendranath et al.,[1] education and engagement of the general public has to be an important component of plans for building capacity to ensure enhanced resilience against natural calamities and extreme events. Citizen science and crowd sourcing have become reliable approaches for timely logistical planning and execution of rescue missions during natural calamities. Citizen scientists, supported with crowd-sourcing tools, were employed for conducting a well-mapping mission after the once-in-a-century floods which hit Kerala state of India in August 2018. The flood affected 5.4 million people, displaced 1.4 million and caused 433 fatalities. Crowd-sourcing was used during the floods for arranging relief camps and enabling victims to request for rescue, food, medical care, food and water supplies and essential sanitary items [2]. A major concern post disaster was assuring access to safe drinking water and sanitation facilities to the public, to avoid disease outbreaks. The floods had disrupted access to public water supply to 6.7 million people, damaged 317,000 shallow wells and nearly 100,000 toilets [3]. Toilets and septic tanks were flooded and overflowed in many areas, enhancing risk of disease outbreaks. There were several cases of acute diarrheal disease (191,945 cases), malaria (518 cases) and chikungunya (34 cases) reported in Kerala in August 2018 [4]. A well-mapping mission was initiated during the floods, aimed at identifying usable wells in areas of selected flood-affected villages, for assessing quality of well water for drinking, and thus reduce the spread of water-associated diseases. Majority of the people in the study area were dependent on public water supply or open wells for their daily water needs and the floods had damaged the water supply systems severely. Therefore, it became necessary to identify the usable wells in the area to ensure the supply of safe drinking water.
The well-mapping mission was conducted by 30 citizen scientists, who visited 300 wells in four days and conducted the study using an online platform with an application downloaded to their mobile phones. Some 37% of the wells in the study area were visually contaminated with floating plants and high turbidity, which might have occurred when flood waters surged over the top of the open wells (Figure-1). Areas surrounding the wells were clean in nearly 70% of the cases (Figure-1), and this was considered by the residents as being most important to avoid the spread of diseases. More than 60% of the wells had septic tank in the proximity (i.e., within 7.5 m) of wells, which indicated high chance of faecal contamination during floods. Local administration and health department had undertaken extensive educational programmes disseminated through social media on the importance of consuming only safe water to avoid water-associated diseases, and all wells were chlorinated multiple times in a week. Since the efficiency of chlorination is largely dependent on the organic load in the well, we graded the wells based on the visual level of contamination and proximity of septic tanks. Those wells with visually clear water and septic tanks at >15 m away were graded as green and suggested for use after chlorination while those with turbid water and septic tanks in the proximity were placed in the red grade and advised to avoid using even for recreational purpose. The wells of different grades were geo-located on an interactive map. Such types of maps are useful for identifying wells of different categories in each area and for preparing a technical plan for their cleaning, frequency of chlorine application, and monitoring.
This work, carried out as part of the Indo-UK project REVIVAL, illustrates how satellite-based communication tools such as smart phones, in combination with citizen science, can provide useful and timely information in the wake of a natural disaster. The work is being extended within the ESA project WIDGEON, in which the use of crowd sourcing and citizen science is being explored, to generate dynamic sanitation maps, which can be updated quickly, in the event of a natural disaster.

References:
[1] Sathyendranath, S.; Abdulaziz, A.; Menon, N.; George, G.; Evers-King, H.; Kulk, G.; Colwell, R.; Jutla, A.; Platt, T., Building Capacity and Resilience Against Diseases Transmitted via Water Under Climate Perturbations and Extreme Weather Stress. In Space Capacity Building in the XXI Century, Ferretti, S., Ed. Springer International Publishing: Cham, 2020; pp 281-298.
[2] Guntha, R.; Rao, S. N.; Shivdas, A., Lessons learned from deploying crowdsourced technology for disaster relief during Kerala floods. Procedia Computer Science 2020, 171, 2410-2419.
[3] Parmar, T.; Manchikanti, S.; Arora, R. Building back better: Kerala addressing post-disaster recovery needs; UNICEF: 2020; p 12.
[4] Shankar, A.; Jagajeedas, D.; Radhakrishnan, M. P.; Paul, M.; Narendrakumar, L.; Suryaletha, K.; Akhila, V. S.; Nair, S. B.; Thomas, S., Elucidation of health risks using metataxonomic and antibiotic resistance profiles of microbes in flood affected waterbodies, Kerala 2018. Journal of Flood Risk Management 2021, 14 (1), e12673.

Europe has the second largest space budget globally and runs a world-class space system with its major pillars Copernicus and Galileo. Copernicus as the European Earth observation (EO) programme is one of the largest data provider in the world with terabytes of EO data generated every day. Both assets Copernicus and Galileo are expected to bring important strategic, social and economic benefits to Europe and the world. In order to ensure the programme exploiting its full benefits, an effective strategy is essential with strong user interaction of the many Copernicus data and information services. The recent and continuing developments in the European space sector stimulate ever-new use cases and applications. Data and services are in place - but their potential is by far not used to its possible extent yet. Evolving needs, taking different stages of a problem-based solution process into account, requires an integrated approach. Apart from the appropriate space technologies and products supplied by service providers, this entails distinctive problem awareness by responsible actors, a workforce with the right skills in place – domain-wise and technical - to address the problem, the dedication to understand and address problems in a joint effort to pull together in the same direction.

The Copernicus User Uptake Programme includes activities to involve proactively stakeholders in the uptake processes of EO-based services, in the adaptation of methods and tools, and in the whole technology and information infrastructure. In the current European Space strategy, the provision of information services and the use of data and policies to promote this is a key element. This applies to a wide range of socially relevant application areas such as environmental protection, transport safety, precision farming, fisheries control, monitoring of shipping routes and detection of oil spills, as well as urban and regional planning. New areas of application are also constantly emerging, including tourism, cultural heritage, supply management or humanitarian aid, to name but a few.

However, experience shows that an integrated approach of demand and supply as well as related skills development [1] is needed to ensure the continuity and sustainability of the evolving downstream sector. Within the Copernicus ecosystem, this involves at least three actor groups: (a) public authorities and bodies as the key beneficiaries of information products and services; (b) strong involvement of the commercial sector as the key providers; (c) a network of academic institutions and champion users, emanating - inter alia – from the Copernicus networks, the Copernicus Academy and the Copernicus Relays. In order to stimulate exchange and interaction between the different actor groups, so-called Copernicus knowledge and innovation hubs shall strengthen the user uptake and development of information services, facilitate an effective transfer of knowledge, encourage cooperation, explore synergies and increase targeted capacity building and training [2]. These hubs can be realised in both ways, as virtual or regional hubs (and blends of each). For the first, main focus is the development of technical elements to visualize and facilitate easy harvesting existing knowledge and experience, while for the latter physically implemented hubs focus on the interaction with local and regional stakeholders and the organization of region-specific implemented outreach and education events.

Within this framework, the University of Salzburg as one of the key player in the Copernicus Academy network organised a sequence of Copernicus-related summer schools, which were also gratefully sponsored by ESA. Entitled “Copernicus for Digital Earth” (2019), “Automated image analysis for the operational service challenge” (2020), and “Intelligent Earth Observation” (2021), those training formats were designed to reach out to different target groups and through this mix of different audiences to cross-fertilise user-driven applications. The summer schools convened participants from students to professionals in the public sector and representatives from companies. Thus, the requirements from authorities were matched with trend-setting R&D originating from academia and solutions tailored by industry.

The most recent international summer school “Intelligent Earth Observation” is a best practice example of integrated approach to close the skills gap between demand and supply in the EO*GI sector workforce, as promoted by the EO4GEO Skills Alliance and. It has pursued a consequent approach of addressing the challenges in a small-scale instantiation of the recently published Sector Skills Strategy . Real-world problems dealt as starting point on which to employ a problem-based application case and build teamwork based solutions. The instructional approach was built upon a combination of the EO4GEO project’s tools and outcomes (Body of Knowledge [3], Curriculum Design Tool, training materials) to establish a compelling training action structure; instructional input was primarily provided by the Skills Alliance partners. Supplemented with keynote inputs from expert guest speakers and embracing a problem based learning approach following the Copernicus downstream idea (i.e., from needs to services), the strongly collaborative summer school can be seen as a best practise example for upskilling and lifelong learning, considering emerging needs of the sector. Particularly oriented towards applications and case-based learning, and following the idea to develop solutions for current challenges, the summer school hosted participants from more than a dozen nationalities who sought for solution building in teams in a virtual environment. The different backgrounds in terms of prior education (bachelor to PhD) and profession (students, researchers, employees in the public and private sector) reflected a welcome mixture of prerequisites to mutually enrich and reinforce the work on application cases. The innovative conceptual design of the Summer School underpinned the strongly case-based learning experience in a three stages: In a first phase, the application areas atmosphere, land and emergency were introduced by means of business and research examples alongside a general introduction, followed by the selection of application cases for group work. The next phase took up EO*GI-concepts useful for group work topics, putting emphasis on artificial intelligence and machine learning. Finally, the third phase was dedicated to elaborate the selected application cases in teams, while consultation and feedback were offered (inquiry-based learning).

References
1. Hofer, B., et al., Complementing the European earth observation and geographic information body of knowledge with a business-oriented perspective. Transactions in GIS, 2020. 24: p. 587–601.
2. Riedler, B., et al., Copernicus knowledge and innovation hubs. International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2020. XLIII-B5-2020: p. 35 - 42.
3. Stelmaszczuk-Górska, M., et al., Body of knowledge for the Earth observation and geoinformation sector - a basis for innovative skills development. International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 2020. XLIII-B5-2020: p. 15-22.

The world is losing 7 million hectares of forests every year, an area that is roughly the size of Portugal. Today, other than the threat of deforestation, we are also dealing with the risks associated with forest fire, pests, diseases, invasive species, drought and other extreme weather events which are putting another 100 million hectares at risk. As the majority of the world’s forests are located in the tropical region, countries within the tropics have set ambitious targets for protecting forests. Their role is seen as critical in reaching the climate goals put forward under the Paris Agreement.
In the last few years, we have seen economic difficulties creep up which have started to impede the efforts put forward by tropical nations, requiring policies and resources to be more effectively aligned. As a result we have seen a rise in public and private commitments to zero deforestation, leading to a more collaborative space in forest governance. Governments that are seeking to reduce greenhouse gas (GHG) emissions by protecting and restoring forests are partnering with private institutions that are motivated to eliminate deforestation through their supply chain. In order for companies to design a sustainable and resilient supply chain, there is an increasing demand for data, especially high resolution satellite data that is affordable and accessible.
An example of such a public-private partnership is the NICFI satellite data program, which is a collaboration between KSAT, Planet and Airbus funded by the Norwegian Ministry of Climate and Environment. Planet and Airbus are providing data for this innovative program that gives free access to high resolution Planetscope monitoring and archival mosaics across the tropical forest region (45 M sq km) for all users as well as historical SPOT5, 6 and 7 scenes over specific areas for selected users. Users have signed up to this program from an amalgamation of backgrounds ranging from governments to NGOs to Journalists, and the Program has partnered with various tools such Global Forest Watch and Google Earth Engine (GEE) to allow for a wider reach of the dataset.

Successful planning and execution of capacity building activities require thorough understanding of the current situation, including detailed awareness of the underlying gaps and opportunities.

Earth Observation (EO) is increasingly used across the globe to support capacity building, in relation to its capability to assist in addressing key economic and societal challenges. To maximise the impact and to increase the efficiency (including resource management) of such activities, decision makers and other actors along the value chain (e.g., research institutes, companies, user communities, investors), require reliable data on the state and progress of different aspects of EO activities in their local EO ecosystem. Assessing Earth observation capacities in such a broad framework is certainly a complex and ambitious task– a diverse variety of factors contribute to the outcome of classifying a country as more or less advanced in the domain. Nonetheless, this complexity does not necessarily justify the fact that currently, reproduceable and comprehensive guidelines as to assessing the maturity of the EO ecosystems at country level do not exist.

The solution we propose in order to fill this gap is the EO maturity indicators (EOMI) methodology. It has been developed and initially implemented under the H2020 GEO-GRADLE project (now a GEO initiative), and has been reviewed and upscaled to its current version for the purposes of the ongoing H2020 e-shape project. The main goal of the EOMI methodology is to create a detailed overview of the country’s EO ecosystem, and thus allow for gap analyses and for identifying strengths. Moreover, a periodic evaluation of a country could particularly aid the assessment of the development of its EO capabilities over time.

In practice, the implementation of the Methodology relies on gathering, assessing, and validating data, in order to attribute one of five maturity levels (0-4) to each of the 49 indicators, distributed in five pillars (Stakeholder ecosystem, Infrastructure, Uptake, Partnerships, Innovation). Vital in this configuration is the role of the “country partner” – usually an institution or private company, and always a player well positioned in the local EO ecosystems to have access to data and validating experts (from fields such as academia, government, industry). The country partner has the leading role in the implementation, supported by the EOMI team – in charge of helping and coordinating the smooth implementation across countries, by providing any support, clarifications, and help – e.g. by offering initial explanations, help identifying national experts to assist the implementation, and continuously reviewing and validating the gathered data. In its full version – the one evaluating 49 indicators across five pillars, the implementation of the EOMI Methodology from beginning to end shall take up to several months. The final outcomes of the implementations are the so called “Maturity cards” allowing for a simplified yet a powerful visualisation of the discoveries of the implementation on country level – by showing the final levels by indicator, as well as grouped by groups and by pillars – allowing to get an initial idea of the gaps and strengths at a glance. Moreover, country partners are encouraged and supported by the EOMI team to publish a more detailed publications of the findings.

The great advantage of the EOMI methodology is its intrinsic modularity and adaptability; each implementing country could in principle choose to only assess some of the proposed pillars or even individual indicators. Moreover, it is possible to adapt the pre-defined indicators and levels to the specificities of the country profile.

Before e-shape, 12 countries have been assessed using the EOMI methodology: 11 in the North Africa, Middle East, Balkan region (under GEO-CREDLE) and \ 1 (independent implementation) in the Philippines. To these countries we add the assessment of EO maturity for 8 European countries under e-shape: Austria, Belgium, Bulgaria, Czechia, Finland, Greece, Italy, Portugal. All these implementations have represented a great opportunity to underline the countries’ strengths and expose their weaknesses. The findings, as well as the details of the methodology, are publicly available, and we highly encourage further uptake.

This paper/presentation will discuss in more detail the needs, the results, and the practicalities of the EOMI methodology, thus showcasing its use and usefulness for assessing EO Maturity at country level, for among others, being able to develop and implement appropriate capacity building activities. In doing so, EOMI can drive investment in future capacities in recognition of identified gaps and opportunities.

This paper deals with the engagement of stakeholders in the context of the new space economy. Therefore it is useful to first introduce the three macro groups of stakeholders usually considered in this context according to the Space Economy Observatory, 2020 (1):
• Upstream stakeholders are "space Industry companies and institutions engaged in research, development, construction and management of enabling space infrastructures and technologies".
• Downstream stakeholders are "companies offering digital innovation solutions and services (e.g., IT provider, system integrator, consulting firm) and specialised research centres that deal with research, development and implementation of the most advanced digital technologies leveraging space technologies and data”.
• End-users are "companies and institutions in demand, interested in new applications and services deriving from the combined use of space and digital technologies.

In a traditional space economy, upstream stakeholders build a satellite constellation commissioned and paid upfront by the client, usually an agency. Thus, the scope, the customers and the envisaged values of a satellite infrastructure are clearly identified since the beginning of the project/programme.
In a new space economy, the liberalisation of the market and the ever-easier access to satellite data have changed the value proposition, particularly for downstream stakeholders and end-users. As an example, the free access to infrastructures such as GNSS has stimulated the emergence of new products, services, businesses and industries. Without the satellite navigation data, downstream such as Google Maps and end-users such as UBER and Deliveroo, it would not exist and, above all, would not be the worldwide giants we all know and have revolutionised mobility. These stakeholders, who extract considerable value, or are even enabled, from satellite data pay negligible amounts for their use, eroding potential revenues for upstream stakeholders. Upstream stakeholders losing potential revenues is the first problem to be addressed.

Furthermore, upstream, downstream, and end-user stakeholders can collect more precise data from many sources. Although data per se are worthless, they should become useful information to stakeholders and thus respond to their needs. On the one hand, upstream stakeholders building satellite infrastructures and sensors, and producing the data, cannot envisage all the usages of their data as they are not end-users experts. On the other end, end-users stakeholders (more and more companies from other sectors such as energy, agriculture, insurance, healthcare) are not aware of the kind of data that satellites might generate and the benefits for their business, as they are not satellite experts. This lack of awareness between upstream and end-users stakeholders and the missed opportunities of exchange value is the second problem to be addressed.

Finally, the complexity and deep uncertainties affecting the medium-long term development of this business may limit the potential of generating value for the society. The main factors to be considered are e.g.
• the heterogeneity of the applications complicates the identification of downstream and end-users stakeholders, their needs and engagement strategies;
• private and public stakeholders may extract value from satellite data; they engage stakeholders in very different ways and with different purposes;
• different stakeholders can access data in different countries;
• the same satellite data can be valuable for different industries and different purposes.
The need to handle this complexity and uncertainty factors is, therefore, the third problem to be addressed.
How can these three problems be addressed? First, to foster the capacity building within the new space economy, we developed a stakeholders engagement framework that helps upstream, downstream and end-users stakeholders to identify strategies to engage. It is intended to be a sensemaking and easy tool to identify the stakeholders and to choose the most suitable engagement approach according to the situation (Figure1).

Given the three categories above of stakeholders, there are three possible links: upstream-downstream, downstream-end-users, end-users-upstream. For each link, a 2x2 matrix represents the domain of stakeholders' engagement in a space project; therefore, there are four main cells in each matrix (Figure 1). Each includes the possible engagement strategies between the two categories of stakeholders. Engagement between the stakeholders may change during the project lifecycle. Introducing the temporal dimension allows us to grasp the dynamics of engagement among stakeholders, according to their characteristics, their relationship in different phases of the project, identifying which engagement strategies to adopt and why they are effective according to the occurrence. Therefore, we decide to investigate the change of the engagement between the same stakeholders in different project phases: "before the beginning of the project", "at the beginning of the project", and "at the end of the project".
The example presented in Figure 2 shows the engagement dynamics among downstream and end-users stakeholders. Each colour corresponds to the engagement between a pair of stakeholders. They are i) Red, engagement between stakeholders A and B; ii) Green, engagement between stakeholders A and C; iii) Blue, engagement between stakeholders B and C. Furthermore, the engagement in three different phases of the project is represented with different symbols. They are i) circle, before the beginning of the project; ii) rhombus, at the beginning of the project; iii) triangle, at the end of the project. The engagement between the pairs of stakeholders in the different phases of the stakeholder project is mapped in the framework, favouring a dynamic vision of engagement (Figure 2).

Let's pretend that the project consists in the development of a remote sensing project to monitor the water leaks of aqueducts; the stakeholders involved are:
• Stakeholder A (end-user): a big private company operating in the Energy sector. Interested in exploring the adoption of remote sensing satellite technologies to monitor the water leaks of their aqueducts. Although the company has understood the potential of satellite technology, they do not know possible suppliers that can provide a solution to their problem, and they cannot engage them. Therefore, they decide to participate in networking events run by experts, designed to bring stakeholders far from the space industry closer to it.
• Stakeholder B (downstream service provider): start-up providing an innovative satellite remote sensing service to monitor the water leaks. The company is not known and still need to establish its reputation. Therefore, they decide to participate in networking events run by experts designed to bring stakeholders far from the space industry closer to it.
• Stakeholder C (downstream data provider): big private company operating in the IT sector. The company acquires, storage, processes the raw satellite data and makes them available and usable on payment to innovative service providers.
The possible engagements are:
• Engagement between stakeholders A (end-user) and B (downstream service provider).
1) Before the beginning of the project, they don't know each other and cannot start engaging. They participate in a networking event run by experts, designed to bring stakeholders together. Here the company representatives get to know each other (Red circle).
2) A, intrigued by the services offered by B, understands that they could be the right service provider to monitor the water leaks of its aqueducts. A starts exchanging emails, phone calls and organise face-to-face meetings with B. This engagement lead to a pilot project using the technologies of B to monitor the water leaks of the aqueducts of A (Red rhombus).
3) During the project, A and B exchange information, knowledge and resources. At the end of the project, B delivers to A a software to monitor water leaks. The project is a success, and the stakeholders' relationship is consolidated; A asks B for an extension of their service (Red triangle).
• Engagement between stakeholders A (end-user) and C (downstream data provider).
1,2,3) In all project phases, they don't know and engage each other. A doesn't have the interest and the capabilities to engage C because they cannot manage raw data. C doesn't consider A as a possible client or partner (Green circle, rhombus and triangle).
• Engagement between stakeholders B (downstream service provider) and C (downstream data provider).
1) Before the beginning of the project, B knows and buy data from C. C provides catalogue data to B without any personalisation, therefore there is no engagement (Blue circle).
2) At the beginning of the project, given the novelty of the application, B engages C to provide a tailor-made set of satellite data. They collaborate together, B shares information about the end user, C invests its resources and technologies to develop a dataset (Blue rhombus).
3) At the end of the project, C includes the new dataset in its product portfolio, which B can access by paying a fee as for the data products requested before the project (Blue triangle).

The engagement between the stakeholders has been influenced by different approaches (e.g., active participation in networking events) with different outcomes. We are building a set of stakeholder engagement approaches for each quadrant of the matrix, assessing which context they are effective (or not) and why. We will also provide a set of guidelines to adopt them. We are looking for stakeholders interested in developing this tool. We are keen to keep participants informed, and we are open to further collaborations.

(1) Permanent research project within the School of Management of Politecnico di Milano

The knowledge of river discharge is crucial for both water resources management activities and for flood risk mitigation. In situ river gauge stations are normally used to monitor river discharge but they suffer from many limitations such as low station density, incomplete temporal coverage as well as delays in data access. Therefore, the development of methods to estimate river discharge from satellite data is strategic especially over data scarce regions where the decline of availability in situ observation data seems inexorable.

In the last decade, ESA has funded different initiatives in the field of discharge estimation, such as the SaTellite based Runoff Evaluation And Mapping and River Discharge Estimation (STREAMRIDE) project, which proposes the combination of two innovative and complementary approaches, STREAM and RIDESAT, for estimating river discharge.
The innovative aspect of the two approaches is an almost exclusive use of satellite data. In particular, precipitation, soil moisture and terrestrial water storage observations are used within a simple and conceptual parsimonious approach (STREAM) to estimate runoff, whereas altimeter and Near InfraRed (NIR) sensors are jointly exploited to derive river discharge within RIDESAT. By modelling different processes that acts at the basin or at local scale, the combination of STREAM and RIDESAT are able to provide less than 3-day temporal resolution river discharge estimates in many large rivers of the world (e.g., Mississippi, Amazon, Danube, Po), where the single approach fails. Indeed, even if both the approaches demonstrated high capability to estimate accurate river discharge at multiple cross sections they are not optimal under certain conditions such as in presence of densely vegetated and mountainous areas or in non-natural basins with high anthropogenic impact (i.e., in basin where the flow is regulated by the presence of dams, reservoirs or floodplains along the river; or in highly irrigated areas).

Here, we present some new advancements of both STREAM and RIDESAT approaches which help to overcome the limitations encountered. In particular, specific modules (e.g., reservoir or irrigation modules for STREAM approach) as well as algorithm retrieval improvements (e.g., to take into account the sediment and the vegetation for RIDESAT algorithm) were implemented. Furthermore, in order to exploit the complementarity of the two approaches, the two river discharge estimates were also integrated within a simple data integration framework and evaluated over sites located on the Amazon and Mississippi river basins. Results demonstrated the added-value of a complementary river discharge estimate with respect to a stand-alone estimate.

Over the last few years, satellite radar altimetry measurements have become more and more numerous over inland waters, thanks to improvements in the tracking function as well as enhanced measurement modes (from pulse-limited low resolution mode to high resolution synthetic aperture radar). This higher quality and abundance of measurements also raises the question of their processing: from historical retracking algorithms (e.g. the Offset Centre Of Gravity - OCOG retracking) to most recent innovative ones (e.g. Fully-Focused SAR). In particular, it is of common knowledge in the hydrology community that current ground segments using the OCOG algorithm provide, on frequent occasions, biased water surface height estimates and is not reliable at global scale. The need for more reliable and robust processing methods is therefore one of the biggest challenges for radar altimetry over land.
The main challenge is that contrary to ocean altimetry, the radar signal over inland waters is highly variable both in space and time, as it depends on the nature and size of the water body observed (lakes, rivers, flood plains, etc.) as well as on surface conditions.
In this study, we focus specifically on rivers where various types of waveforms can be acquired: sinc²-like peak (the most frequent ones), asymmetric peak, multiple peaks, distorted Brown-like waveform. We address the representativeness of the signal’s specularity, as it has been documented in the literature that the radar backscattered signal over rivers is often highly specular and can be modeled as a squared cardinal sine function. We use two parallel approaches: simulation of radar signals over a specular surface, and analysis of real data acquired by current Sentinel-3 and Sentinel-6 Ku-Band missions over rivers.
In the theoretical approach, we use simulations of various specular surfaces and analyze both the amplitude and phase behaviors. It is interesting to analyze the relative impact of the observing system configuration (e.g. radar bandwidth, altitude, sampling) and the nature of the surface (e.g. geometry of the observed scene, backscatter coefficient) on the simulated radar signal.
In the data analysis approach, we use a unique dataset of one hundred rivers worldwide to perform a statistical analysis of the specular nature of the signal and other characteristics in function of the scene configuration. River width is one but not the only parameter impacting the measured signal.
With this study, we aim at understanding the most significant factors controlling the measured radar signal over rivers in order to build a robust and universal processing algorithm capable of providing reliable water surface height estimates to inland waters users.
In the perspective of the Surface Water and Ocean Topography (SWOT) mission, nadir altimetry more than ever stands as an important asset for calibration and validation of this mission as well as for the design of future altimetry missions such as the Copernicus Sentinel-3 Next Generation: Topography mission, which will necessarily address the question of processing performance over inland waters.

The main motivation for this study is to evaluate the use of real-time observations from different sources for hydrological forecasting. The advent of new satellite missions providing high-resolution observations of continental waters has raised the question of how to use them, especially in conjunction with models. At the same time, the multiplication of extreme events such as flash floods points to the need for tools that can help anticipate such disasters. To do so, it is necessary to set up a forecasting system that is generic enough to be used with different types of data and to be applied to different basins. It is in this perspective that a platform named HYdrological Forecasting system with Altimetry Assimilation (HYFAA) was implemented, which encompasses the MGB large-scale hydrological model and an EnKF module that corrects model states and parameters whenever observations are available. As a preliminary study towards operationnability, the platform was tested in offline mode, in the framework of Observing Systems Simulation Experiments (OSSEs). Discharge estimates from three different observing systems were generated, namely in-situ streamflow measurement stations, Hydroweb radar altimetry, and the future SWOT interferometry mission. In this study, we chose to assimilate these data separately in order to analyze the capacity of the system to adapt itself to different orbital characteristics, especially coverage and repetitivity. This also allows us to quantify the contribution of SWOT. The MGB model, developed within the large-scale hydrology research group of the University of Rio Grande do Sul (Brazil), is a physically based and distributed hydrological model, which was coupled to an externalized Ensemble Kalman Filter (EnKF) to give corrected estimates of the model state variables and parameters.
HYFAA is run on the Niger river basin over a reanalysis period and its performance against a control ensemble simulation (without data assimilation) is assessed to quantify the impact of assimilating observations from the different observing systems. The results show that regardless of the observing system, data assimilation generally improves discharge estimation on the basin. We, therefore, discuss limits and perspectives for application in the framework of Observing System Experiments (using real observations).

Satellite altimetry products provide dense observations of water surface elevation (WSE) that can inform hydrodynamic modeling and stage-discharge rating curves estimation. In the present study, we used multi-mission satellite altimetry, e.g., ICESat-1/2, Jason-2/3, and Sentinel-3A/B, to calibrate and validate hydrodynamic models for a 1000-km reach of the Yellow River, the second-longest river in China. We first calibrated spatially distributed Manning-Strickler coefficients by using a steady-state hydraulic solver and ICESat-1/2 altimetry. Considering that river ice jams in the cold season increase channel wetted perimeter, reduce channel hydraulic radius, and increase effective channel roughness, resulting in increased river stages in some river segments, two different scenarios were assumed: river roughness parameters were stable in time (S1); river roughness was varying between the cold and warm seasons, and thus two sets of roughness parameters were calibrated separately (S2). The calibrated sets of roughness parameters were further applied to configure a detailed one-dimensional hydrodynamic model, i.e., MIKE HYDRO River, to simulate time-continuous WSE along the river course. The simulated WSE was validated by satellite altimetry datasets monitored at 24 virtual stations by Jason-2/3 and Sentinel-3 A/B. The calibrated hydrodynamic model was used to determine stage-discharge rating curves for the virtual stations. The discharge of the virtual stations was estimated and validated by comparing with gauging data. The quality and robustness of the rating curves were analyzed. Several factors determine the quality and uniqueness of rating curves (proximity to tributaries, hydraulic hysteresis), and hydraulics at VS locations should be investigated in detail before rating curves are combined with satellite altimetry to estimate discharge.

Space based platforms are an important source of information for the conservation and protection of cultural heritage. The International Council on Monuments and Sites (ICOMOS), a leading international organisation with strong links to cultural heritage conservation utilising space-based resources, will introduce its endeavours.

ICOMOS works to conserve and protect cultural heritage places. It is the only global non-government organisation of this kind, which is dedicated to promoting the application of theory, methodology, and scientific techniques to the conservation of the architectural and archaeological heritage. ICOMOS is a network of 11 000 experts that benefits from the interdisciplinary exchange of its members, among which are architects, historians, archaeologists, art historians, geographers, anthropologists, engineers and town planners. The members of ICOMOS contribute to improving the preservation of heritage, the standards and the techniques for each type of cultural heritage property: buildings, historic cities, cultural landscapes and archaeological sites.

ICOMOS’ International Scientific Committees, partner organisations, associated academic institutions and many of its members are actively utilising data from space-based platforms to undertake research into the impacts of climate change; human activity, ranging from urban development to armed conflict; and to undertake archaeological and other heritage-related research.

Among the main actors is the ICOMOS / ISPRS Committee for Documentation of Cultural Heritage, CIPA Heritage Documentation, an international non-profit organisation that endeavours to transfer technology from the measurement and visualisation sciences to the disciplines of cultural heritage recording, conservation and documentation. CIPA thus acts as a bridge between the producers of heritage documentation and the users of this information. The ability to monitor heritage from space has proved to be a powerful tool in heritage management.

The ICOMOS International Committee on Risk Preparedness (ICORP) enhances the state of preparedness within the heritage institutions and professions in relation to disasters of natural or human origin. It promotes better integration of the protection of heritage structures, sites or areas into national, local as well as international disaster management, including mitigation, preparedness, response and recovery activities. By sharing experience and developing a professional network, ICORP aims to stimulate and support activities by ICOMOS National and International committees for enhancing disaster risk management of cultural heritage. ICORP also supports ICOMOS in its role as the founding partner of the Blue Shield. Data from space-based platforms is an integral aspect of heritage risk preparedness, analysis and response.

This presentation, by a range of experts from ICOMOS, will offer an overview of how space-based monitoring of cultural heritage is now integral to enhancing, better protecting and conserving humanities’ rich and diverse cultural heritage. Scientists, engineers, historians, heritage practitioners, scholars and social scientists are encouraged to attend this session to gain information, establish and enhance their networks, and explore future collaborations.

Background
This work assesses the suitability of a range of satellite imagery for a) detection of buried archaeological and cultural heritage sites; b) monitoring the condition of known archaeological assets; and c) land-use assessment on a national scale. This contributes to a broader ambition to assess remote sensing methods for national-scale survey and heritage management [1]. Over the last couple of decades, satellite data has been reliably used in archaeological site detection and monitoring [2, 3], primarily in arid regions. The climate and weather conditions, land-use and nature of cultural heritage sites in Scotland and Northern Europe make these regions challenging to assess using similar methods [4].
The available satellite data was assessed for frequency of coverage, the ease and accuracy with which proxy indicators of archaeological features (such as crop and soil marks) could be identified, and the ease and accuracy with which land-use changes which have particular implications for cultural heritage sites could be determined. Scotland was used as a trial region, with the understanding that the developed methodology could be applied to other regions with similar climate, land-use and geology.
Satellite data was supplied by the European Space Agency and Planet Labs Inc. Optical data was provided for a variety of regions of interest across Scotland at ground sample distances (GSDs) of 0.5-3m. All imagery was collected in the spring and summer months (April – July) of 2018 and 2020. Particularly dry summers in these years produced high numbers of crop proxies in Scotland. The distribution of data across the spring and summer months was selected to allow for a comparison between imagery to be made over this summer period.

Data Availability
The available data was evaluated to consider the frequency with which suitable data could be reliably acquired of the regions of interest. This process included both automated filtering of the data, and manual examination of the imagery. Data was considered unfit if it did not cover a sufficient proportion of the area of interest, or if it was too cloudy to produce a clear image. Considering all available imagery, approximately 15% of available images were found to be useable. This included imagery that was collected in response to direct tasking requests [5]. Additionally, the usable images were often collected within narrow time periods (e.g. the same or subsequent days) making them unsuitable for prospection methods relying on crop mark identification and limiting their value in both change detection and condition monitoring.

Data Suitability
All the imagery proved suitable for broad brush landscape assessment. Using the data with < 1m GSD an experienced photo-interpreter could visually identify vegetation type and general landscape characteristics. However, the reduced resolution in off-nadir images can make the interpretation challenging. For imagery with > 1m GSD, confidence in interpretation diminishes.
For monitoring the condition of designated ancient monuments, the < 1m GSD imagery provided adequate generalised views of vegetation cover, which can be a good proxy for condition. However, the satellite imagery, even at its best in a nadir view, does not have the resolving power required to identify specific items of concern, such as discrete areas of rabbit burrowing.
For identification of monuments, the < 1m GSD data has proved to be adequate for detecting variation in vegetation down to c. 1m across. However, with many of the archaeological features commonly found in Scotland often being ≤1m, the resolution of the imagery puts it on the cusp of reliably resolving such features.

Conclusions
The available satellite data has the potential for use in historic environment applications including land-use assessment, condition monitoring, and archaeological prospection. Datasets with < 1m GSD are of significantly higher value compared to data with > 1m GSD; the reliability and confidence with which the latter could be used for all proposed applications was notably lower than the former. However, the intermittent availability of suitable imagery is a significant limitation. Although the presented work focuses on Scotland as a case study region, it is expected that the outcomes would apply across regions of similar climate across Europe. Were < 1m GSD satellite data available to the European historic environment community with suitable frequency and reliability, it could be of significant value.

References
[1] Banaszek, Ł., Cowley, D. C., & Middleton, M. (2018). Towards national archaeological mapping. Assessing source data and methodology—A case study from Scotland. Geosciences, 8(8), 272. doi: https:// doi.org/10.3390/geosciences8080272
[2] Lasaponara, R., & Masini, N. (2011). Satellite remote sensing in archaeology: past, present and future perspectives. Journal of Archaeological Science, 9(38), 1995-2002. doi: 10.1016/j.jas.2011.02.002
[3] De Laet, V., Paulissen, E., & Waelkens, M. (2007). Methods for the extraction of archaeological features from very high-resolution Ikonos-2 remote sensing imagery, Hisar (southwest Turkey). Journal of Archaeological Science, 34(5), 830-841. doi 10.1016/j.jas.2006.09.013
[4] Cowley, D. C. (2016). Creating the cropmark archaeological record in East Lothian, southeast Scotland. Prehistory without Borders: Prehistoric Archaeology of the Tyne-Forth Region; Crellin, R., Fowler, C., Tipping, R., Eds, 59-70.
[5] McGrath, C. N., Scott, C., Cowley, D., & Macdonald, M. (2020). Towards a satellite system for archaeology? Simulation of an optical satellite mission with ideal spatial and temporal resolution, illustrated by a case study in Scotland. Remote Sensing, 12(24), 4100. doi: 10.3390/rs12244100

The region of eastern South Africa plays a major role in the development of the technical and behavioural capacities, that allowed Homo sapiens to expand into Europe and Eurasia during the Late Pleistocene. Although the region yields some of the best-studied archaeological sequences, it is still understudied in terms of site density and open-air sites, which could help to better understand the connectivity and use of space of ancient populations. We present two applications of remote sensing for archaeological prospection, that are tailored for the specific detection of (a) open-air sites and (b) rock shelters.
Colluvial landforms are suitable archives as they preserve archaeological artefacts bedded in sediment layers that provide context information on past landscape stability, climate and vegetation changes and allow radiometric dating methods. We mapped these landscape features through the use of multispectral remote sensing and digital landscape analysis. The spectral properties of local surface and soil profile materials were characterized in situ through field spectroscopy (250-2500 nm wavelength), yielding high resolution reflectance curves that give insight to their physio-chemical properties. Thereupon we developed spectral indices, which enable the discrimination of different surface types and applied these to VIS, NIR and SWIR bands of WorldView-3 to map the colluvia based on its specific spectral properties.
Rock shelters host most of the currently known archaeological sites of the region, where excellent preservation conditions allow to study millennia old anthropogenic remains with resolutions of centuries and even decades. We predicted the presence of potential sites through the analysis of a high accuracy DEM (TanDEM-X). The application of geomorphometric and hydromorphometric indices together with stratigraphical information allowed to derive the specific properties of this landform.
Our results show a large number of yet unidentified potential archaeological sites of both types. They lay a foundation for a DFG-funded project starting 2021 (WI 4978/3-1), that will evaluate the results in a interdisciplinary framework of researches from domains like archaeology, geography, geology, chronometry, paleoproteomics, biogeology, etc.

Over the last two decades remote sensing of satellite imagery for cultural and natural heritage (CNH) risk assessment has significantly increased. Corroborating evidence of such a growth of interest is found in the scientific publications, white papers, policy documents and more generally in the grey literature. At the same time, protection and safeguarding of cultural and natural heritage have been raised higher in international agendas (e.g. as a target in the UN SDG #11) and what satellite technologies can specifically do towards this scope is the subject of reflections at least at European level (e.g. Copernicus Cultural Heritage Task Force).
Despite these major advances, the majority of studies focused on selected types of damages such as looting, natural-hazards and conflicts-related that were generally deemed as the most dangerous ones (Zaina 2019) or that were more consistently covered by the increased flow of information and media attention in response to the events occurring in different regions. This narrative has been recently tackled by studies asking for a more comprehensive understanding of the entire set of damage to CNH. In particular, they shed light on other equally or more dangerous but under-considered threats including ploughing, construction of roads and buildings or even large infrastructures. Among the latter, the construction of dams resulted in the flooding of thousands of archaeological sites in different places of the world. Despite this pervasive effect, specific regulations, as well as tailor-made methodologies to document and monitor this type of damage, are yet to be codified (Marchetti et al. 2020).
Current counteractions are spotty and suffer from a lack of coordination by international and national authorities, while support from ad hoc legislations is also missing in many countries. Moreover, mitigation strategies – e.g. activities of rescue archaeology that the present paper aims to target as a specific application domain for which satellite data may be a useful support for archaeologists in areas of dam construction – mostly rely upon ground-truthing activities and, in some cases, aerial imagery. The limits of these strategies include: 1) incomplete research methodology, which does not consider the potential of remote sensing for sites identification; 2) limited timeframe, as ground-truthing allows only to identify the latest types of damages that are visible at the time of in-situ surveys; 3) incomplete geographic coverage, as confirmed by many archaeological surveys in prospective reservoir dams; 4) low level of detail, as ground-truthing does not allow to appreciate anomalies visible for space.
These shortcomings could be effectively overcome by integrating remote sensing of satellite imagery. It is, in fact, well-known that archaeological research methodologies are highly benefitting from the growing availability of open-access satellite imagery and their accessibility through different online platforms like Google Earth and Bing Maps (Agapiou 2017), thus opening new research avenues, including their application for archaeological damage assessment and monitoring. However, what open data from Copernicus Programme, as well as licensed data from Contributing and Third Party Missions, can do in the context of dams construction has not been fully explored yet. Furthermore, the use of satellite data is not yet an established practice across the archaeology community.

In this context, this paper aims at showing the potential of multidisciplinary collaboration between image analysts and archaeologists to carry out activities of site documentation from remote for rescue archaeology, by which Copernicus Programme Sentinels and Contributing Missions data are integrated to assess the impact of dams on cultural heritage. Building from the three protocols system proposed by Marchetti et al. (2020), we focus on how satellite imagery can input into the first protocol, named Pre-Construction Archaeological Risk Assessment (PCARA), consisting in the quantification of archaeological evidence located within the prospective reservoir area before dam construction and/or when impoundment/reservoir filling takes place. The integrated methodology encompasses the following steps:
1. Reconstruct dam construction/impoundment timeline and identify user needs (e.g. documentation prior to flooding and loss vs. assessment of residual risk in case partial damage or loss already happened);
2. Search for baseline data to check whether inventories of sites at potential risk are already available or an ex-novo site survey is required;
3. Search for archive satellite imagery matching the temporal framework of dam construction and identify user requirements for new imagery collection to perform change detection and time series analysis;
4. Task satellites accordingly or make an informed selection of routinely collected imagery, also accounting for variables on the ground that may impact the quality of the observations;
5. Manual surveying conducted by archaeologists with expertise in site identification and good background in remote sensing (the latter being an advantageous skill, that may be also built during the collaboration and support from the image analysts);
6. Site mapping and risk assessment, with iterative process back to satellite data selection and/or new image acquisition repeated, better or refined satellite observations are needed.
Ideally, this workflow should be completed with validation in the field through ground-truthing, that is however beyond the scope of this presentation.
We showcase the feasibility of this methodology on two case studies, the small planned dam of Halabyeh in Syria and the large, currently under construction Grand Ethiopian Renaissance dam.
The first case is an arid region poorly covered by archive satellite imagery. Therefore, besides the use of already available open access dataset, a robust acquisition strategy of imagery was required to integrate the dataset. To this aim, we tuned the satellite acquisition schedule by programming the full Italian Space Agency’s COSMO-SkyMed constellation to collect high-resolution SAR images to meet the needs of archaeological research.
In the second case, we chose a humid region well documented by both SAR (COSMO-SkyMed) and multispectral imagery (Sentinel-2) prior to the reservoir filling. The temporal acquisitions collected over this area by different satellites cover both low and high vegetative periods.
One of the main lessons learnt specifically regards step 4. In order to achieve the most accurate site identification, it is essential to consider the following two types of variables that can influence the detectability of an archaeological site: 1) temporal; 2) environmental.
The temporal variable relates to the timeframe of the dam construction and how satellite data can provide sufficient temporal coverage of the different phases (i.e. pre-, during and post-construction). Therefore, this variable influences the selection of the archive and new images to be acquired. In particular, in the case of an area with few archive images, a new acquisition campaign has to be carried out.
The environmental variable regards the physical properties of the natural setting where the dam is constructed that affect the visibility of archaeological features in the satellite data. Above all, seasonality. Indeed, highly vegetative landscapes may limit visibility. Therefore, if the prospective reservoir area is located in an arid or semi-arid region, seasonality may not be a constraint for satellite images collection. On the contrary, humid regions may require specific satellite data tasking during low or high vegetative periods.
A careful analysis of these variables allows to efficiently tune selection of archive satellite data and collection of new ones to meet end-users (from academic researchers to private and public) specific needs. In this regard, a valuable support has been recently provided by Synthetic Aperture Radar (SAR) and multispectral imagery of the different space agencies’ constellations (Tapete and Cigna 2019). These open-access and licensed images integrate each other providing both landscape (multispectral imagery) and small scale archaeological feature (SAR) identification, global spatial coverage, high temporal revisit and ease of data handling.
Another important result is represented by the number of variables useful to identify CNH that emerged from the manual identification based on COSMO-SkyMed and Sentinel-2 derived products. These include: 1) the shape of the archaeological sites; 2) the terrain colour; 3) unexpected irregularities of some elements of the territory; 4) the location of anomalies along abandoned meanders of the rivers considered.
We aspire that these achievements will pave the way for the integration of the workflow in a detailed PCARA protocol to be used in the development of ad hoc policy papers for international and national stakeholders to protect CNH threatened by dams.
For the purposes of ESA LPS22, this paper will address several objectives of session D2.12, such as, highlighting the benefits of multidisciplinary collaborations and partnerships between different heritage experts, demonstrating how to shape the exploitation of Sentinel and other Copernicus missions data to meet end-users needs, and contributing to the evidence base via sharing of developed use-cases and respective lessons learnt.


REFERENCES
Agapiou, Athos. 2017. “Remote Sensing Heritage in a Petabyte-Scale: Satellite Data and Heritage Earth Engine© Applications.” International Journal of Digital Earth 10 (1): 85–102.
Marchetti, Nicolò, Gabriele Bitelli, Francesca Franci, and Federico Zaina. 2020. “Archaeology and Dams in Southeastern Turkey: Post-Flooding Damage Assessment and Safeguarding Strategies on Cultural Heritage.” Journal of Mediterranean Archaeology 33 (1): 29–54.
Tapete, Deodato, and Francesca Cigna. 2019. “Detection of Archaeological Looting from Space: Methods, Achievements and Challenges.” Remote Sensing 11 (20).
Zaina, Federico. 2019. “A Risk Assessment for Cultural Heritage in Southern Iraq: Framing Drivers, Threats and Actions Affecting Archaeological Sites.” Conservation and Management of Archaeological Sites 21 (3): 184–206.

The European Directive on open data and the re-use of public sector information (2019; a successor of PSI Directive, 2003) identifies “Earth observation and environment” as one of high-value datasets categories that is to say “documents the re-use of which is associated with important benefits for society, the environment and the economy, in particular because of their suitability for the creation of value-added services, applications and new, high-quality and decent jobs, and of the number of potential beneficiaries of the value-added services and applications based on those datasets”. On a global scale, Group for Earth Observation (GEO) has launched a dedicated Initiative, EO for Sustainable Development (EO4SDG) in Service of Agenda 2030 (Global Goals for Sustainable Development).
Within such a framework of open geospatial information (including EO) in support to specific SDGs (namely Goal 11 “Sustainable cities and communities” and goal 15 “Life on land”), this paper explores how satellite earth observation (EO) for purposes of cultural heritage monitoring is being portrayed across Europe, within both scientific and grey literature.
The objective of this study is to answer several research questions: (i) firstly, what are the main categories of damages to cultural heritage studied using satellite imagery in Europe so far (scientific literature)? Is it possible to assess a group of the most studied sites until now and the threats that have been associated to them? And, in case, what (if any) impact have these studies produced on the work of public administrations and site managers, and end-users in general (grey literature) on several identified case studies?
In order to explore the above listed aspects, the authors have taken into consideration that over the last few decades, the long-term trend of risks and damages to the archaeological heritage has received an increasing interest among scholars. Building upon a previously tested methodology (Zaina & Cuca, 2022), the paper is looking to identify an exhaustive number of documents that provide a focus on the evolution of the indexed scientific literature and grey literature (reports, working and government documents, white papers and so on) relying on the use of EO for monitoring of Cultural Heritage and Cultural Landscapes.

Few hundred papers regarding the sites across Europe published since 2005 have been identified for the purpose, following the procedure described below:
• Firstly (Step 1), papers were automatically retrieved from the Scopus® scientific database using several different combinations of synonyms for “damage” (relating to satellite, heritage and archaeology), which are present in the title, abstract and keywords of each indexed paper. We chose Scopus because it is the largest abstract and citation database of peer-reviewed literature including thousands of scientific journals, books and conference proceedings and it delivers a comprehensive overview of the world's research output in any field of study.
• From the resulting sample of 1531 papers, we applied the Excel® “automatic duplicate removal” to delete 837 duplicate papers (Step 2).
• The remaining 694 papers were then subjected to a detailed manual screening (Step 3) in order to exclude all those which, although including two or more keywords, did not deal specifically with the topic of damage to cultural heritage. The manual screening was carried out by analyzing the title, abstract and keywords and, in case of poor or lack of exhaustive information, by directly accessing the paper. Examples of off-topic papers are those analyzing climate change or catastrophic events (e.g., earthquakes, volcanic eruptions) that occurred in the past.

Hence a total number of almost 700 on-topic papers emerged from the manual screening and are subject to further investigation.

In parallel, a second survey has been carried out to screen and investigate the grey literature, under the following motivations:
• the scientific literature alone cannot provide an exhaustive representation of demonstration activities involving or made directly by users and stakeholders (as observed in previous studies; e.g. Tapete & Cigna, 2017), given that journal papers are mostly focused around applied research, methodological developments and tests, as well as proof of concept or case studies, thus it is sometimes unfeasible to grasp the real impact of EO technologies on daily practice;
• not all demonstration activities have necessarily translated into papers or be presented at conferences with indexed publications, whereas reports from European projects and/or practitioner associations are more informative in this regard;
policy, programmatic and institutional documents issued by organisations and bodies in charge of heritage preservation can provide insights into the status of EO technology uptake and embedding in operational workflows.

Based on the analysis of such diverse documentations, the discussion will be around:
• apparently uneven distribution of evidence base across topics of concern for heritage documentation and preservation, in relation to the fact that some of these topics are directly related to specific functions (e.g. inventorying sites and regular monitoring of their condition) and/or the existence of regulatory frameworks, national or European policies or directives (e.g. the European Landscape Convention, Common Agricultural Policy) that the heritage administrations are requested to undertake and address;
• the variety of situations that seems to be found across Europe, from (i) heritage administrations that are already acquainted with satellite technologies (not rarely as a recent extension from consolidated expertise in aerial photography) so as to include them among their data sources and means for documentation, (ii) to situations where the technology is yet to be approached or the demonstration process has not yet been followed by a local capacity built or established;
• the role of “facilitator” / “accelerator” that scientific partners or specialist consultants can play to help heritage administrations to take advantage of EO technologies;
• the benefit from multidisciplinary collaboration, especially in governmental and international initiatives.

The main envisaged contribution of this analysis is a potential betterment of procedures in cultural heritage sites protection, monitoring and, ultimately, management. The results of this work aim to further contribute to reframing priorities for recording, documenting and monitoring of built heritage, especially bigger sites inserted in a larger framework of Cultural Landscapes.

The human presence in the Arctic has resulted in a rich legacy of archaeological, historical and cultural sites, many of which are at risk from anthropogenic impacts. Climate change in the Arctic impacts human heritage on a devastating scale, including coastal erosion, damaging or destroying coastal archaeological sites, to biological and chemical processes that are accelerating, causing decay, disturbance and, ultimately, the destruction of objects and structures. Climate change has also provided opportunities for the theft of woolly mammoth tusks from previously frozen ground, many from archaeological sites, and increased general tourist 'souveniring' of material culture.

Local communities cultural safety is at risk through the theft, damage or destruction of cultural objects and places of spiritual, cultural and historical significance. Central to the research is respectful engagement with the Peoples of the North to coproduce knowledge that is requested by them. There will be a strong emphasis on providing communities with opportunities and support for knowledge and transfer of skills in how to utilise remote sensing resources.

Space-based resources are already providing data that has increased knowledge of littoral and terrestrial changes, and changes in flora and fauna and enhanced responses by communities, archaeologists and heritage practitioners. Increasingly, indigenous peoples have utilised space-based resources to produce data for use in their endeavours to retain their traditional lifeways and enhance their lifestyles.

New possibilities for assessment and mapping of known sites, discovery of previously unrecognised sites, three-dimensional mapping, change-detection and site monitoring at high temporal resolution are emerging as a result of a number of developments in both spaceborne and airborne remote sensing technology and associated geoinformatic approaches. Especially significant developments are the maturing of structure-from-motion techniques, the rapid development of publicly accessible cloud-based image processing coupled with an increase in availability of medium-resolution satellite imagery, the expansion in scope and sophistication of free open-source software, the growing overlap between scientifically capable and consumer-grade UAV systems, and changes in the regulations relating to the use of UAVs in many countries. These changes can all potentially further increase the scope for co-production of knowledge and understanding of indigenious and other cultural assets; enhancing cultural heritage understanding, appreciation, conservation and protection.

Scientists, engineers, polar historians, heritage scholars and other social scientists are encouraged to attend this session to gain information, establish and enhance their networks, and explore future collaborations.

The CAScade Smallsat and Ionospheric Polar Explorer (CASSIOPE) spacecraft containing the enhanced Polar Outflow Probe (e-POP) instrument suite was launched in 2013 by the Canadian Space Agency. In 2018, the spacecraft joined the European Space Agency (ESA) Swarm virtual constellation through ESA’s Third Party Mission program as Swarm-Echo. Here we present the new Swarm-Echo magnetic field data product based on the Magnetic Field instrument (MGF). The Swarm-Echo dataset follows the format and usage conventions of the Swarm Level 1B product definitions as much as possible, though, owing to differences in the platforms and sensors, some adjustments are necessary. Swarm-Echo carries two fluxgate magnetometers but lacks an absolute scalar field sensor. The Swarm-Echo data product is calibrated in-situ based on a vector-vector comparison to the CHAOS magnetic field model that has been enabled by a significantly improved attitude solution. We present algorithm used to create this dataset, the criteria used to select data suitable for calibration, and implementation of the linear inverse algorithm used to solve for the twelve standard magnetometer calibration parameters. Prior to calibrating the sensors, the data are corrected for transients that occur when update the magnetic feedback in the instrument. Then, we reduce the data to one sample per second, and bin it into seven-day intervals. The data are then culled using information from the attitude, bus telemetry, and location files, as well as conditions given by the Kp and Disturbance storm time (Dst) indices. For calibration we select data that falls within ± 55° latitude during geomagnetically quiet times. We consider geomagnetically quiet to be when the Kp index does not exceed 3 or the Dst index does not exceed a change of 3 nT per hour when the data were taken. From the attitude files, we flag any data where the attitude solution was not generated by at least one of the star tracker cameras due to the large (up to 30°) error in solutions generated from the CSS as mentioned in section 4. We also flag any data where the solution has dropped out for greater than ten seconds or there is greater than ten seconds until the next signal is obtained due to potentially large errors when interpolating the attitude solution. Lastly, we flag any data where the angular rotation rate of the spacecraft exceeds 0.03 degrees/sec. The new data product typically has residuals from CHAOS below 10 nT during geomagnetically quiet times and is now operationally produced and distributed for use by the scientific community.

Enlarging datasets of magnetic space-based observations of the Earth’s magnetic field is of interest to space physics and geomagnetism. E.g., geodetic satellites carrying platform magnetometers are suitable missions as they fly in low earth and in near polar orbits. Filling gaps of data of high-precision magnetic missions, such as was the case between CHAMP and Swarm, as well as an increased spatio-temporal coverage make the use of satellite’s platform magnetometer data attractive.
We calibrate and characterize platform magnetometer data of different geodetic satellite missions (GOCE, GRACE-FO, and others) by machine learning techniques. By applying machine learning techniques, the measured magnetometer signal is adapted for artificial disturbances from other satellite payload systems. The proposed non-linear regression can automatically identify relevant features as well as their crosstalk, allowing for a wider range of available inputs. This approach reduces analytical work for the calibration of platform magnetometers, thus leading to faster, more precise, and easily accessible magnetic datasets from non-dedicated missions. The calibrated datasets are made publicly available.
With the use of as much information as possible about the satellite’s payload systems’ activations as possible, our approach models the magnetic behavior in-situ and post-launch. To this aim, data on the temperature, currents as well as telemetric data is collected and incorporated to automatically find relevant inputs for the magnetic behavior of the satellite as a system. In addition, the proposed method automatically identifies arbitrary time shifts in the inputs, e.g., in measurements of the platform magnetometers. In this process, the CHAOS-7 model acts as the reference model which is backed by high-precision (Swarm magnetic) data as well as ground observatory data.
The evaluation shows promising results, reducing the error compared to the reference model significantly. Results of the characterized and calibrated magnetic datasets by machine learning techniques are evaluated by demonstrating enhanced representations of ionospheric and magnetospheric currents and lithospheric signatures. It shows that appropriate handling of multi-satellite magnetic data can increase the geomagnetic datasets significantly in parallel to only one high-precision mission.

Observations from the Swarm mission have made possible imaging of the magnetic field generated inside Earth’s core to unprecedented detail. These data, along with physics-based assumptions concerning the dynamics of the liquid iron within the core, permit us to construct new data-constrained models of the direction and speed of flow at the edge of the Earth’s core. Such maps of the flow lie at the heart of understanding how the Earth’s magnetic dynamo operates some 3000km beneath the surface.

The unique constellation configuration of Swarm’s polar orbiting satellites has provided new insights into magnetic field change at high latitude. Of particular interest is the behaviour of a cluster of localised patches of magnetic field at the core surface, located at around 70 degrees latitude, which serve as a powerful indicator of core dynamics. The movement and change in structure of these patches was recently explained by an accelerating jet (Livermore et al., 2017), inferred to be localised on the region of core surface at high-latitude stretching beneath Siberia to Canada. Such behaviour of the flow has not yet been identified elsewhere. We will present new images and models of the dynamics of this phenomenon which is apparently unique to the north polar region, based on the most recent data from Swarm and ground-based observatories. This work comprises one of the goals of the Swarm 4D-Earth Core ESA-funded consortium.


The identification of localised core flow acceleration at high latitude has profound implications for our understanding of the large-scale dynamics of the core, as this high-latitude region forms an important component of the core’s general circulation pattern which consists of an planetary-scale eccentric gyre (Pais & Jault, 2008). This gyre connects, on the one extreme, strong westward-directed flows at high latitude, with, on the other, inferred westward directed flows under the Atlantic on the equator. A better understanding of the global pattern of flow within Earth’s core requires not only continued space-based monitoring of high latitude magnetic field by Swarm, but also an improved understanding of equatorial core dynamics from the monitoring of low-latitude internal magnetic field that will be made possible by new geomagnetic missions with improved local time coverage, such as the proposed NanoMagSat mission.

Livermore, P. W., Hollerbach, R., & Finlay, C. C. (2017). An accelerating high-latitude jet in Earth’s core. Nature Geoscience, 10(1), 62–68.

Pais, M. A., & Jault, D. (2008). Quasi-geostrophic flows responsible for the secular variation of the Earth's magnetic field. Geophysical Journal International, 173(2), 421–443.

Detailed mapping of the Earth's magnetic field brings key constraints on the composition,
dynamics, and history of the crust. Satellite and near-surface measurements are complementary as they detect different length scales.
We present a magnetic field model built after a selection and processing of magnetic field measurements from the German CHAMP and ESA Swarm satellites, which we merge with a world compilation of near-surface scalar anomaly data. The global model is constructed from a series of independent regional models that are then transformed into a unique set of spherical harmonic (SH) Gauss coefficients. This procedure relies on parallelization and memory optimization and allows us to generate the first global model to SH degree 1050 derived by inversion of all available measurements from near surface to satellite altitudes.
The model agrees with previous satellite-based models at large wavelengths and fits the CHAMP and Swarm satellite data down to expected noise levels. Further assessment in the geographical and spectral domains shows that the model is stable when downward continued to the Earth's surface. However, we observe a possible reduction of the expected power spectrum between SH degrees 120 and 200 that arises from a low signal to noise ratio in the available satellite and near surface data. Numerical simulations demonstrate that the measurements of the magnetic field scalar and vector gradients at steadily decreasing altitudes during the Swarm A and C satellites re-entry will provide invaluable low altitude measurements for filling this gap. Additional simulations carried out in the framework of the NanoMagSat project, which aims to deploy a new constellation of three identical nanosatellites, using two inclined (approximately 60°) and one polar Low Earth Orbiting satellites, show that the Earth’s lithospheric field will also be better depicted in the spectral range.

The solar activity in the form of coronal mass ejections leads to abnormal geomagnetic field fluctuations. These fluctuations, in their turn, generate so-called geomagnetically induced currents (GIC) in electric power grids, which may pose a significant risk to the reliability and durability of such infrastructure. Hence, nowcasting and ultimately forecasting GIC is one of the grand challenges of modern space weather studies. One of the critical components of such now/forecasting is a real-time simulation of the ground electric field (GEF), which depends on the electrical conductivity distribution inside the Earth and the spatiotemporal structure of geomagnetic field fluctuations. In this contribution, we present a methodology that allows researchers to simulate the GEF in a real time, irrespective of the complexity of the Earth's conductivity and geomagnetic field fluctuations models. The methodology relies on a factorization of the source by spatial modes and time series of respective expansion coefficients and exploits precomputed frequency-domain GEF generated by corresponding spatial modes.

The validation of the presented concept is performed using Fennoscandia as a test region. The choice of Fennoscandia is motivated by several reasons. First, it is a high latitude region where the GEF is expected to be particularly large. Second, there exists a 3-D ground electrical conductivity model of the region. Third, the regional magnetometer network, IMAGE, allows us to build a realistic source model for a given geomagnetic disturbance using the Spherical Elementary Current Systems (SECS) method. Factorization of the SECS-recovered source is then performed using the principal component analysis. Finally, taking several geomagnetic storms as space weather events, we show that the real-time high-resolution 3-D modelling of the GEF at a given time instant on a 512x512 lateral grid is feasible and requires only a fraction of a second provided that frequency-domain GEF due to the preselected spatial modes are computed in advance.

We also discuss how one could implement the presented methodology for nowcasting and forecasting the GEF using ground-based magnetic field data, ACE/DSCOVR satellite observations of the solar wind parameters at the L1 Lagrangian point, and magnetic field data from the low-Earth-orbit active geomagnetic missions, like Swarm and CSES, and planned missions like Macao and NanoMagSat.

Heliophysics, the science of understanding the Sun and its interaction with the Earth and the solar system, has a large and active international community, with significant expertise and heritage in the European Space Agency and Europe. Several ESA directorates have activities directly connected with this topic, including ongoing and/ or planned missions and instrumentation, comprising a ESA Heliophysics observatory. A particularly strong example of this has been the cross-directorate collaboration between Swarm and Cluster. Discussions in this area began in the early stages of Swarm, shortly after selection and grew into funded activities examining synergies between the missions, including dedicated ISSI working groups and a forum. Both missions continue to operate with evolving scientific targets, which include specific cross- mission science, only possible through such collaboration. This paper will provide a brief overview of Cluster -Swarm activities, past present and future.

Methane (CH4) is the second most important anthropogenic greenhouse gas (GHG) in terms of its overall effect on climate radiative forcing. The atmospheric residence time of methane is considerably shorter than that of carbon dioxide, but its warming potential significantly stronger. Methane is produced from natural sources such as wetlands, and as a result of human activities, such as the oil and gas industry. A small number of anomalously large anthropogenic point sources are a major contribution to the total global anthropogenic methane emission budget, thus early detection of such sources has great potential for climate mitigation.

Satellite observations allow the detection and quantification of methane sources even in remote areas where monitoring would otherwise be difficult or costly. Methane satellite observations are now possible from a wide variety of instruments with very high spatial resolution. In this work, we explore the capabilities of three satellites with different specifications and spatial resolutions ranging from metres (WorldView-3; multi-spectral) to tens of metres (PRISMA; hyperspectral) to kilometres (TROPOMI; hyperspectral), to detect and quantify methane point sources.

We use TROPOMI Level 2 XCH4 data from IUP Bremen and statistical methods to find areas with methane anomalies in selected regions of the world. We then use PRISMA and WorldView-3 (WV-3) observations to zoom in those areas, and identify and quantify methane emissions from small point sources.

For PRISMA and WV-3 we use a fast data-driven retrieval algorithm to derive the methane column enhancements. We then use an automated system based on a deep learning model to detect and isolate methane plumes. This model has been trained using synthetic methane plumes generated with the Large Eddy Simulation extension of the Weather Research and Forecasting model (WRF-LES), and embedded in the WV-3 or PRISMA radiances. Finally, we calculate the emission flux rates using the well-known Integrated Mass Enhancement (IME) method.

We present several case studies of observations of methane plumes from the three satellites, and characterise their performance. This work has been done by researchers from the National Centre for Earth Observation based at the University of Leicester, University of Leeds and University of Edinburgh as part of a project funded by the UK Natural Environment Research Council (NERC).

The importance of reducing methane emissions to mitigate climate change in the short term has been recognised at the highest political level. At COP26 over 100 countries signed up to the Methane Pledge to reduce methane emissions by 30% by 2030. The International Methane Emission Observatory (IMEO) was recently launched by the UN to support companies and governments in reducing their methane emissions by providing actionable data based on a.o. methane measurements. In addition, the European Commission has prominently identified the role of satellites to support its goal to reduce methane emissions. As such TROPOMI -in combination with other satellites- can play an important role to support these efforts.

The Tropospheric Monitoring Instrument (TROPOMI) aboard ESA’s Sentinel-5P satellite was launched in 2017 and provides daily global coverage of methane concentrations at up to 7 x 5.5 km2 resolution. These data can be used to detect persistent methane emissions as well as large emission events such as accidents in the natural gas industry. As TROPOMI provides 10-20 million observations a day, we use a machine learning technique to detect the methane super-emitter signals. This machine learning approach allows constant global monitoring. While some plumes can be attributed to a specific source using just TROPOMI data, its resolution usually does not enable that. Therefore, we use the TROPOMI super-emitter detections to guide high-resolution instruments such as GHGSat, PRISMA, and Sentinel-2 to pinpoint the exact source(s) responsible, information needed for mitigation. For persistent sources, we use long-term TROPOMI data combined with metrological data to pinpoint the source’s location as precisely as possible. The meter-scale observations of high-resolution instruments over limited domains can then be used to identify the exact facilities responsible for the enhancements seen in TROPOMI and estimate their emissions. In this way we have been able to identify super-emitting landfills, oil/gas facilities, and coal mines. This information can be used to inform the operators allowing – for example – gas leaks to be fixed. The satellite-based emission estimates can also be used to evaluate reported emissions. We will show the latest results from this synergistic use of these different satellites.

Combining TROPOMI with the high spatial resolution satellites is a very powerful tool to not only understand but also mitigate anthropogenic methane emissions. As such we have already communicated the exact locations of quite a number of methane super emitters to the IMEO for further action.

The Paris Agreement foresees to establish a transparency framework that builds upon inventory-based national greenhouse gas emission reports, complemented by independent emission estimates derived from atmospheric measurements through inverse modelling. The capability of such a Monitoring and Verification Support (MVS) capacity to constrain fossil fuel emissions to a sufficient extent has not yet been assessed. The CO2 Monitoring Mission (CO2M), 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 present a CCFFDAS that operates at the resolution of the CO2M sensor, i.e. 2km by 2km, over a 200 km by 200 km region around Berlin. It combines models of sectorial fossil fuel CO2 emissions and biospheric fluxes with the Community Multiscale Air Quality model (coupled to a model of the plume rise from large power plants) as observation operator for XCO2 and tropospheric column NO2 measurements. Inflow from the domain boundaries is treated as extra unknown to be solved for by the CCFFDAS, which also includes prior information on the process model parameters. We discuss the sensitivities (Jacobian matrix) of simulated XCO2 and NO2 troposheric columns with respect to a) emissions from power plants, b) emissions from the surface and c) the lateral inflow and quantify the respective contributions to the observed signal. The Jacobian representation of the complete modelling chain allows us to evaluate data sets of simulated random and systematic CO2M errors in terms of posterior uncertainties in sectorial fossil fuel emissions. We provide assessments of XCO2 alone and in combination with NO2 on the posterior uncertainty in sectorial fossil fuel emissions for two 1-day study periods, one in winter and one in summer. We quantify the added value of the observations for emissions at a single point, at the 2km by 2km scale, at the scale of Berlin districts, and for Berlin and further cities in our domain. This means the assessments include temporal and spatial scales typically not covered by inventories. Further, we quantify the effect of better information of atmospheric aerosol, provided by a multi-angular polarimeter (MAP) onboard CO2M, on the posterior uncertainties.

The assessments differentiate the fossil fuel CO2 emissions into two sectors, an energy generation sector (power plants) and the complement, which we call “other sector”. We find that XCO2 measurements alone provide a powerful constraint on emissions from larger power plants and a constraint on emissions from the other sector that increases when aggregated to larger spatial scales. The MAP improves the impact of the CO2M measurements for all power plants and for the other sector on all spatial scales. Over our study domain, the impact of the MAP is particularly high in winter. NO2 measurements provide a powerful additional constraint on the emissions from power plants and from the other sector. We explore the effect of the random component in the error of the NO2 measurements on the posterior uncertainty of inferred fossil fuel emissions.

The United Nations Paris Agreement requires countries to reduce their greenhouse gas emissions. Transparent monitoring of CO2 emission reductions is desirable at multiple spatial scales, including the scale of individual facilities such as fossil fuel burning power plants. Here, we present a case study on monitoring CO2 emissions from Poland’s Bełchatów coal-burning power plant (the highest CO2 emitting power plant in Europe), using space-based observations from NASA’s Orbiting Carbon Observatory 2 and 3 (OCO-2 and OCO-3) missions. In addition to standard observations from these missions, we also use the ‘Target’ observing mode from OCO-2 and ‘Snapshot Area Mapping (SAM)’ mode of OCO-3. The emission estimates are compared with reported power generation, demonstrating the ability to detect emission changes over time at the facility scale. This research gives us a glimpse of the potential capabilities of future CO2 imaging missions like the Copernicus Anthropogenic CO2 Monitoring Mission (CO2M) and reinforces the value of future space-based CO2 data for verifying local-scale CO2 emission reductions with the potential to support the transparency framework under the Paris Agreement.

Accurate and up-to-date greenhouse gas (GHG) emission inventories are essential in developing targeted policies designed to reach net zero targets. Efficiently processing the large volumes of data being produced by satellites allows us to detect changes in anthropogenic emissions of GHGs. Satellite observations are an integral part of up to date emission reporting and monitoring. Directly relating GHG observations to emission hotspots is non-trivial due to the long life times of GHGs and the influence of natural components on the CO2 and methane fluxes. By analysing co-emitted species from combustion, we can determine the major emission sources of GHGs. We have developed a unique analysis tool, using a convolutional neural network (CNN), to identify plumes of nitrogen dioxide (NO2), a tracer of combustion, from NO2 column data collected by the TROPOspheric Monitoring Instrument (TROPOMI). Our approach allows us to exploit the growing volume of satellite data available to determine the locations of emission hotspots around the globe on a daily time scale. We train the deep learning model using six thousand 28 x 28-pixel images of TROPOMI data (corresponding to ~266 x 133 km2) to find emission plumes of various shapes and sizes. The model can identify plumes with a success rate of more than 90%. Over our study period (July 2018 to June 2020), we detect over 310,000 individual NO2 plumes, each with a corresponding location and timestamp. We can relate these locations to known emission hotspots such as cities, power stations and oil and gas production with over 9% of the detected plumes located over China. Ship tracks through the Suez Canal, South East Asia and around The Cape of Good Hope also become visible within the processed data. Our tool shows mixed results when compared to gas flaring regions, highlighting the difficulty of creating an accurate dataset for this type of emission. We have attempted to remove the influence of open biomass burning using correlative high-resolution thermal infrared data from the Visible Infrared Imaging Radiometer Suite (VIIRS). We also find persistent NO2 plumes from regions where inventories do not currently include emissions, including mid-Africa and Siberia, demonstrating the potential of this tool to help update inventories. Using an established anthropogenic CO2 emission inventory (ODIAC), we find that our NO2 plume distribution captures 92 % of total CO2 emissions, with the remaining 8 % mostly due to a large number of sources smaller than 0.2 gC/m2/day for which our NO2 plume model is less sensitive. We believe our type of analysis, used in conjunction with other earth observation products, can form part of a suit of products to improve estimates of anthropogenic emissions of greenhouse gases.

Space-based measurements of carbon dioxide (CO2) are the backbone of the global and national-scale carbon monitoring systems that are currently being developed to support and verify greenhouse gas emission reduction measures. Current and planned satellite missions such as JAXA’s GOSAT and NASA’s OCO series and the European Copernicus Anthropogenic Carbon Dioxide Monitoring (CO2M) mission aim to constrain national and regional-scale emissions down to scales of urban agglomerations and large point sources with emissions in excess of ~10 MtCO2/year.

We report on the demonstrator mission “CO2Image”, now in Phase B, which is planned for launch in 2026. The mission will complement the suite of planned CO2 sensors by zooming in on facility-scale emissions, detecting and quantifying emissions from point sources as small as 1 MtCO2/year. A fleet of CO2Image sensors would be able to monitor nearly 90% of the CO2 emissions from coal-fired power plants worldwide. The key feature of the mission is a target region approach, covering tiles of ~50x50km2 extent with a resolution of 50x50m2. Thus, CO2Image will be able to resolve plumes from individual localized sources, essentially providing super-resolution nests for survey missions such as CO2M.

Here, we present the instrument concept which is based on a spaceborne push-broom imaging grating spectrometer measuring spectra of reflected solar radiation in the SWIR-2 spectral window around 2µm. It relies on a comparatively simple spectral setup with a single spectral window and a moderate spectral resolution of approximately ~ 1 nm. The instrument is designed to fly in a sun-synchronous orbit at an altitude of 500 to 600 km. We further discuss the overall mission goals and evaluate the mission concept in terms of e.g. optimal local overpass time and sampling strategy.

The mass loss of the Greenland Ice Sheet is substantially affected by surface melt , but the amount of liquid water that stored form in the firn layer is not well constrained. While the interior of the ice sheet is only occasionally experiencing surface melt , the ablationzone is prone to intensive melting over the summer months . In highly crevassed shear zones and other regions of high stress changes firn layers are even more accessible to melt water.
In soil, ponding of water over cracks in summer leads to distinct infiltration characteristics. We assume that this also applies for firn, and therefore investigate whether areas of ponding over crevasses are locations where radar images indicate the retention of water. To this end we will use different SAR missions in X, C, and L-band, with a focus on L-band.

Our study is focused on the area of 79°N Glacier, an outlet glacier of the North East Greenland Ice Stream. We conduct a Pauli decomposition of ALOS-1 and ALOS-2 data to distinguish between different scattering mechanism. We find features of very low backscatter in all polarizations, that are moving at the speed of the glacier. Furthermore, we use data acquired with AWI’s ultrawideband (UWB) radar from 2018 and 2021 to investigate the internal structure in the upper hundred meters of the glacier across such features. We compare our findings from observations with simulated firn/ice transition depth using an Lagrangian along-track approach. In addition to that, we classify crevasse zones and aim to constrain their depth with the ice penetrating radar. This way, we can assess whether water infiltration in summer is the origin of the observed features.

It is essential to advance our knowledge of the hydrology of the Greenland Ice Sheet, to assess its current contribution to global sea-level rise, and to improve its projection into a future warming climate. The high latitudes of the Northern Hemisphere have experienced a large regional warming over the last decades, and as a consequence the amount of liquid water at the surface of the Greenland Ice Sheet has significantly increased.
The 4DGreenland project (https://4dgreenland.eo4cryo.dk/) is funded by the ESA as part of the POLAR+ programme. It runs for two years (2020-2022) and its main objective is to maximize the use of Earth Observation (EO) to assess and quantify the hydrology of the Greenland ice sheet. The project is focused on dynamic variations in the hydrological components of the ice sheet, and on quantifying the water fluxes between the hydrological reservoirs.
Here, we present the initial results and data products, and analysis from the project. These include:
• improved mapping of drainage and refilling of active subglacial lakes using different satellite sensors
• basin-wide subglacial melt from an EO-informed model on a monthly resolution
• meltwater volume contained in supraglacial lakes based on optical imagery and ICESat-2 derived lake bathymetry
• identification of surface melt extent from SAR and PMW, with value added products such as melt onset, duration and intensity.
We have used a wide range of satellite data sets, and have developed, improved, implemented and tested methods to map the different hydrological components from these. We have combined the EO-derived data products with output from regional climate models to generate a basin-scale monthly time series of runoff from the Greenland ice sheet for selected test sites. In the next phases of the project, the analysis will be carried out ice sheet wide and for the time period 2010-present (the CryoSat-2 period). All products will be available at (as a minimum) monthly resolution. Some products will be available for a longer time period and at a higher temporal resolution.

As climate warming intensifies the melting of the Greenland Ice Sheet, increased fluxes of meltwater drain from its surface to the ocean. The passage of water on top, through, and ultimately underneath the ice is important, because of the influence it exerts upon future ice sheet evolution. This critical, but elusive, system is being studied using a suite of satellite data by ESA’s Polar+ 4D Greenland study. Here, we report analysis of ~35,000 high-resolution (2 metre) time-stamped Digital Surface Models (DSMs), to search for evidence of previously unknown hydrological phenomena around the entire margin of the Greenland Ice Sheet. This new dataset provides unprecedented detail of localised surface elevation changes, and allows us to identify previously unknown hydrological signals.

In this presentation, we focus on a detailed case study of a large – and highly unusual – subglacial lake drainage event identified beneath the Brikkerne Glacier, in northern Greenland. In summer 2014, approximately 9x107 m3 of water drained from this previously unknown subglacial lake, causing an 80 metre drop in the ice surface. Further downstream, the resulting subglacial flood breached back through the surface of the ice sheet, causing fracturing of the ice sheet surface, and depositing ice debris measuring 25-metres in height. To our knowledge this is the first time that such behaviour has been observed in an ice sheet setting, and reveals a complex, bidirectional coupling between the surface and basal hydrological systems. Analysis of historical satellite imagery suggests that the subglacial lake drained previously, in 1990; however, on that occasion the subglacial floodwater failed to breach the ice surface. This suggests that climate warming and ice sheet thinning over the past three decades may have helped to facilitate the passage of subglacial floodwater to the surface. As such, our observations demonstrate a new, and poorly understood, aspect of Greenland hydrology and show, more broadly, how high-resolution streams of satellite data can be used to obtain insight into the hydrological system. This work forms a contribution to ESA’s Polar+ 4D Greenland study.

The ice velocity of the Greenland Ice Sheet can be monitored from space using data from the ESA Sentinel-1 mission. The PROMICE ice velocity product exploits the high temporal and spatial resolution of the Sentinel-1 SAR data to produce Greenland Ice Sheet wide ice velocity mosaics that resolves variations on the seasonal scale. The product is a continuously updated time series (September 2016 - present) of mosaics produced using intensity offset tracking of Sentinel-1 SAR data and is posted on a 500 m grid. Each mosaic spans two Sentinel-1 A cycles (24 days) and a new one is made available every 12 days, and is provided within approximately 10 days of the last acquisition.

In this contribution, we use machine learning to investigate the seasonal dynamics of the Greenland Ice Sheet as well as the interannual variations. The length, coverage and the high temporal and spatial resolution of the time series make it possible to study not only a few locations on the ice sheet, but to investigate the flow of ice over large parts of the ice sheet and over several years.

We perform a fully automated, unsupervised clustering of annual time series in all grid points in the fast flowing part of the Greenland Ice Sheet. The analysis reveals how a number of well defined clusters describe the general seasonal flow patterns. It also shows how the clusters dominate different regions of the ice sheet and different areas of the outlet glaciers. The patterns of where different clusters dominate are not stationary but show significant interannual variations. By correlating to other data sets e.g. modelled surface run-off and topography, we can infer information on the processes linked to the different seasonal behaviours.

We identify spatial patterns in the seasonal velocity along all the marine outlet glaciers of the Greenland ice sheet and show that the patterns depend on the local availability of meltwater and vary upstream from the glacier front. Our findings represent a significant step forward in understanding the processes controlling the marine outlet glaciers compared to previous studies that were restricted to include a limited number of datapoints and only considered a fraction of the outlet glaciers. Our results underline the importance of obtaining long time series of ice velocity covering the entire Greenland ice sheet and in high spatial and temporal resolution, and provide insights needed to improve estimates of the future dynamic mass loss from the Greenland ice sheet.

Near-surface density is a key parameter for assessing the overall mass balance of ice sheets. Specifically, it is used during the conversion of surface elevation changes inferred from laser and/or radar altimetry into mass changes. As such, constraining the full mass balance of an ice sheet requires insight of spatiotemporal patterns in near-surface density that themselves are typically derived from global and/or regional climate models. While these models are calibrated with a handful of in situ measurements, there is currently no observational approach for constraining large-scale spatial and temporal patterns in near-surface density. In face of a changing climate, a measurement-driven approach to understanding of the spatial and temporal variability in surface density would represent a major step forward in reducing the uncertainty in estimates of ice sheet dynamics as well as future sea-level rise.

Beginning with ERS-1 in the early 1990’s, conventional space-borne radar altimetry studies of ice sheets have been overwhelmingly focused on using the time delay between signal transmission and the reception the reflected echo to determine how far the spacecraft is from the surface. Repeating these measurements through time and space then allows for ice sheet volume to be monitored across spatiotemporal scales unattainable using either airborne or in situ methods. At the same time however, little attention has been paid to the strength with which radar waves are reflected from the Earth’s surface. This is not unsurprising as measured radar wave reflection strengths are affected by several physical factors including, but not limited to, the contrast in electromagnetic properties across the Earth’s surface interface as well as the roughness of that interface.

The desire to quantitatively assess radar surface echo strengths represents a unique opportunity to introduce new methodologies initially developed for other planetary surfaces into the analysis of Earth Observation datasets. Specifically in this study we investigate using the Radar Statistical Reconnaissance (RSR) method originally developed for and applied on Mars. RSR leverages the statistical distribution of radar echo strengths over an assumed to be homogeneous region in order to separate the coherent and incoherent components of the reflected echo. The coherent component is related to the contrast in electromagnetic properties across the surface interface while the incoherent component is a function of surface roughness. Therefore, by applying the RSR methodology to surface echo powers recorded within a region, we can begin to quantitatively describe the geophysical parameters of that area.

In this work, we present early results from applying the RSR methodology to Ku- and Ka-band satellite radar altimetry measurements of the Greenland Ice Sheet (GrIS). We make use of minimally processed waveforms recorded by the SIRAL (Ku-band) instrument on-board ESA’s CryoSat-2 satellite as well as the ALtiKa (Ka-band) instrument on-board the joint ISRO/CNES SARAL satellite. To maximize surface echo density, and therefore RSR spatial resolution, we have focused our CryoSat-2 efforts thus far on the analysis of SARIn waveforms covering the marginal portions of the GrIS as opposed to less-dense LRM waveforms which are restricted to the interior. We generate RSR results for two instruments with different inherent range resolutions (SIRAL and ALtiKa signal bandwidths of 320 and 500 MHz respectively) in order to assess any impact of near-surface structure on our interpretations.

In addition to clear patterns in both coherent and incoherent powers, the RSR results reveal unique features dependent on data source. As SIRAL SARIn waveforms are acquired at a much higher rate (bursts of 64 pulses at 17.8 kHz, with a burst repetition frequency of 85 Hz) than those from ALtiKa (pulse repetition frequency of 40 Hz), the maximum radial distance around each RSR grid point required to gather a relevant statistical population of surface echo powers is much greater for ALtiKa as opposed to SIRAL. It then follows that for complex surfaces, such as those near the ice sheet margin, because surface echoes from a larger area must be included in the SARAL/ALtiKa RSR analysis, the likelihood that the echo powers reflect a homogenous surface is reduced. This stationarity assumption is implicit in RSR, which when violated, degrades the reliability of the ALtiKa results in these areas as opposed to SIRAL. In contrast, as SARIn measurements are performed only along the margins of the GrIS, the reliability of the current CryoSat-2/SIRAL RSR results diminishes towards the interior.

Moving forward, in order to incorporate the RSR results into future GrIS mass balance estimates, the coherent component must be calibrated to reflect near-surface density. Calibration will be performed using the SUMup database, which contains a comprehensive record of near-surface density measurements from firn/ice cores and snow pits from across the GrIS. To achieve maximum compatibility, calibration will be performed using only RSR results from the same months as the in situ density measurements were collected before being applied across the entirety of the RSR dataset.

In closing, by expanding the scope of how radar satellite altimetry datasets are analyzed, this research endeavors to provide new quantifiable and observational insights into the mass balance of the GrIS that serve to help reduce uncertainty with regards to estimating its contribution to future sea-level rise.

Runoff from the Greenland Ice Sheet has increased over recent decades affecting global sea level, regional ocean circulation, and coastal marine ecosystems. Runoff now accounts for most of Greenland’s contemporary mass imbalance, driving a decline in its net surface mass balance as the regional climate has warmed. Although automatic weather stations provide point measurements of surface mass balance components, and satellite observations have been used to monitor trends in the extent of surface melting, regional climate models have been the principal source of ice sheet wide estimates of runoff. To date however, the potential of satellite altimetry to directly monitor ice sheet surface mass balance has yet to be exploited. Here, as part of the Polar+ Earth Observation for Surface Mass Balance (EO4SMB) project, we explore the feasibility of measuring ice sheet surface mass balance from space by using CryoSat-2 satellite altimetry to produce direct measurements of Greenland’s runoff variability, based on seasonal changes in the ice sheet’s surface elevation. Between 2011 and 2020, Greenland’s ablation zone thinned on average by 1.4 ± 0.4 m each summer and thickened by 0.9 ± 0.4 m each winter. By adjusting for the steady-state divergence of ice, we estimate that runoff was 357 ± 58 Gt/yr on average – in close agreement with regional climate model simulations (root mean square difference of 47 to 60 Gt/yr). As well as being 21 % higher between 2011 and 2020 than over the preceding three decades, runoff is now also 60 % more variable from year-to-year as a consequence of large-scale fluctuations in atmospheric circulation. In total, the ice sheet lost 3571 ± 182 Gt of ice through runoff over the 10 year survey period, with record-breaking losses of 527 ± 56 Gt/yr first in 2012 and then 496 ± 53 Gt/yr in 2019. Because this variability is not captured in global climate model simulations, our satellite record of runoff should help to refine them and improve confidence in their projections.

There are ~100 million lakes in the world at a wide range of spatial scales. For large lakes we can derive water quality parameters using ocean colour satellite sensors, which benefit from a long history of algorithm development and validation, and dedicated bio-optical data collection. For the optically complex waters found in lakes a rich body of work has been dedicated to developing algorithms for the MERIS sensor, and the follow-on mission OLCI benefits from this legacy. Due to the optical diversity of inland waters, algorithms intended for global use require in situ validation data from a wide range of optical water types. For the MERIS and OLCI sensors their long time series means a diverse selection of in situ data are more readily available.

The 300m resolution of MERIS/OLCI imposes a lower size limit on the lakes that we can monitor with these sensors, which has led researchers to look to the MSI sensor on Sentinel 2 as a higher resolution alternative. MSI was primarily designed for land remote sensing so lacks the radiometric resolution and sensitivity of the dedicated ocean colour sensors. Furthermore, the relatively short time series limits the availability of in situ data for validation. Currently the available in situ data for MSI does not adequately represent the diversity and seasonality of optical properties of lakes to calibrate and validate algorithms to be used on the global scale.

While our ultimate goal is to improve algorithm performance for MSI using in situ data, we can make improvements in the short term by comparing with established satellite sensors and algorithms. Here, we present work to align the results of turbidity and chlorophyll-a algorithms for MSI against concurrent OLCI data, resulting in significant performance improvements compared with the original algorithm coefficients.

In this presentation we will outline our approach and show results from the tuned algorithms. The remaining challenge is to explore how well these aligned algorithms perform when applied to smaller lakes where issues such as land adjacency affect a large proportion of the water body. We will present examples where the aligned algorithms perform well and where they fail. As the output of these algorithms are available at 100-m resolution in the Copernicus Global Land Service Lake Water Quality product, we will make recommendations on the type of applications and water bodies for which the current procedures are adequate, and on which priorities we see for the further development of high-resolution inland water quality products.

Remote sensing reflectance (Rrs) is the conventional measurement used in aquatic remote sensing to spectrally characterize optically active constituents (OACs) just below the water surface and is often the desired parameter to be used in empirical and analytical bio-optical models for water quality applications. Deriving Rrs from Earth observing multi-spectral image pixels is a major challenge as the water-leaving reflectance is only a fraction of the total signal received by a space-borne sensor due to the interfering scattering and absorbing properties in the atmosphere. Though several algorithms and software have been developed to alleviate this process through atmospheric correction for global ocean color sensors, it is still a very active area of research for medium resolution sensors for inland water remote sensing. To address this concern, the U.S. Geological Survey (USGS) released the Landsat-8 Collection 1 Level-2 Provisional Aquatic Reflectance (AR) Science Product in April 2020 after the Operational Land Imager (OLI) sensor showed a high degree of fidelity to derive reliable Rrs measurements at 30 meters resolution over coastal and inland validation platforms and coincident observations with other ocean color sensors.

The AR science product is based on the Sea-viewing Wide Field-of-View Sensor (SeaWiFS) Data Analysis System (SeaDAS) originally developed by the National Aeronautics and Space Administration Ocean Biology Processing Group (NASA/OBPG) modified for the Landsat-8 OLI sensor. The SeaDAS algorithm processes Landsat-8 Level-1 TOA reflectance bands to estimate Rrs through precomputed radiative transfer simulations that depend only on the sensor spectral response, solar and sensor viewing geometry, and ancillary information such as atmospheric gas concentrations, surface windspeeds, and surface pressure. Spectral Rrs are then normalized by the Bidirectional Reflectance Distribution Function (BRDF) of a perfectly reflecting Lambertian surface (multiplied by π) to produce the dimensionless Aquatic Reflectance. Global AR products for Landsat-8 Collection 1 and Collection 2 data are made available for download through the Earth Resources Observation and Science (EROS) Center Science Processing Architecture (ESPA) on-demand interface to help support ongoing contributions to aquatic remote sensing and environmental monitoring capabilities.

With scheduled release of Landsat Collection 2 AR provisional products in early 2022, improvements were made to the water pixel classification for non-water masking that will enhance the capability to evaluate spectra from AR products across a range of optically diverse targets. For qualitative uncertainty in aquatic reflectance spectra, SeaDAS generates a Level-2 processing flags (l2_flags) band that provides additional per-pixel information about the Landsat-derived spectra. Additionally, intermediate Rayleigh-corrected reflectance (ρrc) products for the visible to shortwave-infrared (VSWIR) bands will be included in the Collection 2 AR product package as well as the original atmospheric auxiliary input raster bands used in the SeaDAS processing for advanced pixel analysis.

As new Earth observing satellite sensors become operational as part of long-term continuity missions designed and launched by NASA (National Aeronautics and Space Administration), there will be a heightened need for data synergy and cross-sensor harmonization to provide increased observational frequency that yield comparable measurements. With the recent launch of Landsat-9 OLI-2 as well as proposed plans for future Landsat missions, the AR science product is a first-step toward structuring a standardized unit of measurement for the health and changing dynamics of aquatic ecosystems.

This presentation will provide the aquatic remote sensing community with a description and general use of the Landsat Provisional Aquatic Reflectance Science Product, its characteristics, product packaging, download accessibility, and future implementation to facilitate its continued use in water management practices and global water surveying.

Soda lakes in the East African Rift Valley are some of the most productive aquatic ecosystems on earth. These lakes sustain more than half of the global population of Lesser Flamingos, which feed on dense blooms of the cyanobacteria Arthrospira fusiformis and microphytobenthos. Given their dependence on a small network of Rift Valley lakes, Lesser Flamingos are highly vulnerable to both climate change and catchment degradation. This study aimed to investigate: i) the extent to which the water quality of soda lakes is changing; ii) what is driving these changes - e.g., land-use or climate change; and iii) how these water quality changes are affecting the numbers and distribution of Lesser Flamingos.

We used Landsat 7 ETM+ to examine the water quality trends of 14 Rift Valley lakes over a 21-year time period. Low cloud images were selected, and monthly composites were generated using the median reflectance for each pixel. A decision tree-based classification algorithm, originally developed by Tebbs et al. (2015) for Rift Valley lakes, was adapted to include salt crust as an additional class. The classification scheme utilises the distinct spectral signatures of different ecologically relevant optical water types to group pixels into pre-defined classes. The final classification included two classes that represent valuable food sources for flamingos (high biomass waters and microphytobenthos) and five further classes (low biomass, sediment dominated, surface scum, bleached scum and salt crust). The method was then applied to the monthly Landsat ETM+ composites to estimate lake ecological status and flamingo food availability.

Environmental parameters were also estimated from satellite imagery to assess the impacts of land-use and climate change on lake ecological states. Using Landsat 7 ETM+ imagery, the Modified Normalised Difference Water Index (MNDWI) was applied to estimate lake water levels and water surface temperature was estimated using the statistical mono-window (SMW) algorithm applied to Top of Atmosphere (TOA) imagery. The Normalised Difference Vegetation Index (NDVI) was applied to MODIS imagery to assess land degradation within catchments and precipitation was estimated using the Climate Hazards Group Infrared Precipitation with Station (CHIRPS) data. All remote sensing analyses were performed in Google Earth Engine. Ground-counts for Lesser Flamingo numbers were obtained from the International Waterbird Census. Water quality (as an indicator of food availability) and flamingo distributions were then modelled using generalised linear mixed models.

Results confirmed that Lake Bogoria and Lake Nakuru in Kenya are important feeding lakes for Lesser Flamingos. Food resources in these lakes have been declining in recent years, without increasing sufficiently at other lakes. Declines in food availability were linked to rising water levels, which are occurring due to a combination of increased rainfall and land degradation within the lake catchments. Food availability was a key driver of flamingo distributions, with water levels also having an influence. These findings have important implications for conservation and land and water resource management within the East African Rift Valley. This study also demonstrates the potential of optical classification of inland waters using Landsat imagery for the long-term monitoring of lake water quality and ecological status.

Water quality is one of the major issues that societies are already facing and there is a strong need of a global monitoring on water quality in inland waters. Satellite data are increasingly viewed as a trustworthy solution to complete the ground observations that are severely lacking in most parts of the world.

Research teams from the IRD (GET Laboratory), OFB and INRAE (ECLA research group) research institutions, supported by the French Space Agency (CNES) have developed the OBS2CO processing chain for inland water color analysis and water quality mapping based on Sentinel-2 images which resolution makes possible to assess very small objects, of a few tenths of meters, such as rivers and minor lakes. The processing chain handles Sentinel-2 Level-1C products and takes advantage of two modules specifically designed for complex optical water and small objects : 1) Glint Removal for Sentinel-2 images that operates both atmospheric correction and sunglint correction (Harmel et al. 2018); 2) Water Detect module that retrieve water pixels using unsupervised images and proved to be very efficient for small water objects, down to 1 ha (Cordeiro et al. 2021). Both codes were tested during large intercomparisons tests (ACIX-AQUA for atmospheric correction over optically complex waters and WorldWater Round Robin for water surface detection) and proved to provide robust results. The OBS2CO chain delivers water occurrence maps as well as water quality maps for different parameters : suspended particulate matter (SPM), turbidity, chlorophyll-a concentration, Harmfull Algal Blooms index, Coloured Dissolved Organic Matter (CDOM) absorption coefficient at 440 nm. A large database of water quality and field radiometric measurements (i.e. > 1000 points) collected in river and lakes across the world is used to evaluate the retrieval algorithms.

The first water color products from Sentinel-2 images are distributed through UNESCO World Water Quality Portal over Lake Chad Region in Africa (https://lakechad.waterqualitymonitor.unesco.org/portal/) and will be shortly available at the French land datacenter. We will explain the method, calibration of the model and its validation over various sites worldwide and present the freely distributed water color products.

The Sentinel-2 high resolution makes possible to retrieve very fine details related to hydrological and biophysical processes in rivers and lakes. Different applications of the OBS2CO processing chain will be presented over Europe, South America, Africa and Asia. These applications consider suspended sediment flux quantification over rivers and reservoirs, eutrophication monitoring in semi arid areas and for EU water framework directive assessment, dissolved organic matter mapping in tropical areas and disaster assessment.

Invasive floating aquatic plants, such as Water Hyacinth (Pontederia crassipes) and Water Lettuce (Pistia stratiotes), have substantial negative ecological and socio-economic impacts, affecting lake water quality and impeding fishing and navigation. Due to their free-floating nature and rapid proliferation, they pose a significant logistical issue in their removal and require continuous ground-based monitoring, which is both expensive and time consuming. Existing methods for remote sensing of floating plants have limited geographic or temporal coverage and are not easily transferable to other open water systems. The main limitations were the spatial, temporal or spectral characteristics of past sensors, now mostly resolved by the recent Copernicus Sentinel-1 and Sentinel-2 missions.

In this presentation, we propose generalised and well-validated methods for satellite remote sensing of floating aquatic plants using radar (Sentinel-1 SAR) or optical (Sentinel-2 MSI) imagery and evaluate the strengths, limitations and complementarity of radar and optical imagery. The spectral properties of floating macrophytes are investigated, and we evaluate spectral indices suited to shoreline delineation and distinguishing floating plants from surface algae and open water. The methods developed were tested on eight diverse waterbodies distributed globally across three continents. We discuss the challenges of ground-truthing floating macrophyte maps due to the dynamic nature of free-floating plants and present a standard protocol for generating validation data from available satellite imagery.

Owing to the similar spectral and backscatter properties of land and floating plants, multi-temporal imagery was used for shoreline delineation. The Automated Water Extraction Index (AWEI) and Normalised Difference Moisture Index (NDMI) applied to Sentinel-2 optical imagery were used to distinguish floating plants, surface algal blooms and open water, and radar imagery from Sentinel-1 VH backscatter was used to distinguish floating plants from open water. Methods were able to map floating plants with user’s accuracies of 86.4% (Sentinel-1) and 79.8 % (Sentinel-2) and producer’s accuracies of 84.3% (Sentinel-1) and 80.4% (Sentinel-2). Timeseries for the eight waterbodies were produced showing the change in floating plant coverage from December 2018 – June 2021.

The strengths of Sentinel-1 radar included being unimpeded by cloud cover and floating surface algae. It was, however, influenced by false positives due to wind and thermal noise artifacts and was less able to resolve narrow river channels. The two satellites provided complementary information on floating weed location and coverage, with significant agreement on same-day measures of floating weed area (R2 = 0.85, p < 0.05). We therefore suggest the benefits of using both sensors together for detection and monitoring of floating macrophytes. Our novel methodological framework provides a valuable tool for researchers and water managers for monitoring the spread of water hyacinth and other floating weeds to target control measures more cost-effectively.

Cyanobacteria occurrences are more and more frequent in many water bodies around the world. Due to their potential harmfulness, the detection and monitoring of cyanobacteria blooms are especially relevant in freshwater bodies for communities with implanted recreational activities as well as for government instances assessing and monitoring water quality for reporting purposes.
Various studies have shown that satellite data can be used to quantify cyanobacteria blooms in lakes and coastal waters. These approaches include the assessment of floating cyanobacteria or the detection of phycocyanin in cyanobacteria blooms. Using remote sensing to detect cyanobacteria in the water column is based on the spectral behavior of phycocyanin, which, in contrast to chlorophyll-a, has a clearer absorption maximum at 620nm. At the same time there is an absorption minimum at 650 nm and at 700 nm. To be able to detect these absorption features with an optical sensor and to distinguish them from chlorophyll-a, the corresponding sensor must have narrow recording channels at the necessary wavelength ranges. The required phycocyanin absorption band at 620nm is only available for a few sensors, such as MERIS (on board ENVISAT 2002-2012) and OLCI (Sentinel-3, since 2016). The spatial resolution of both sensors is 300 m and is therefore only suitable for larger lakes. In the case of sensors with a higher spatial resolution, which are often optimized for land applications, there are usually wider recording channels and the area around 620 nm is not explicitly covered with its own band at 620 nm. Developing a method to detect cyanobacteria with high spatial resolution sensors is therefore not straightforward but essential to not only take smaller lakes into account during monitoring but also to investigate the spatial extent and behavior of cyanobacterial blooms in water bodies. Sentinel-2 MSI has proven to be a valuable sensor for monitoring smaller (inland) waters bodies but is missing the 620nm band. Nonetheless, analyzing these inland water bodies with Sentinel-2 MSI showed that high cyanobacteria abundances present distinctive features in the spectra and the water colour which can be used to derive an indicator for potential risk of cyanobacteria occurrences, by e.g. using a B5/B4 band ratio.
We present in this approach a random forest (RF) model as a regressor to identify the risk of cyanobacteria blooms with Sentinel-2. RF is a regressor that randomly produces multiple decision trees, with each tree representing a class prediction. The class with the most votes becomes the model’s prediction. The RF method is used for regression or classification approaches, and it belongs to the supervised machine learning algorithms. The RF method uses the wisdom-of-crowds concept and is becoming popular in remote sensing applications thanks to easy and fast model training. The importance of the used variables has been extensively tested for different scenarios. A manually selected training dataset with known cyanobacteria blooms, high chlorophyll biomass blooms and clear water cases has been collected covering various inland waters from Germany and the USA. Since atmospheric corrections may fail in extreme blooming events or generate higher uncertainties, we decided to use bottom of Rayleigh reflectances (BRR) spectra instead. This has also the advantage to be independent on a certain atmospheric correction algorithm. Four spectral indices were determined to represent the occurrence of cyanobacteria in inland waters. During the validation phase challenges for the model were identified in brown waters with cyanobacteria occurrences. These challenges were resolved by introducing a second model trained with a subset of the entire training’s dataset. The identification of cyanobacteria is finally based on the combination of these two models. The output of the combined model is a risk status of potential cyanobacteria occurrences. The risk status is subdivided into low, medium, and high risk.
The validation of the developed approach was performed on German lakes, based on an extensive in-situ dataset from the German state offices. For the validation dataset, Cyanobacteria occurrences from in-situ data were classified if the biovolume of cyanobacteria were higher than half of the total biovolume. Additionally, a minimum chlorophyll-a concentration of 10µg/L needed to be recorded to account for a cyanobacteria bloom. Furthermore, a selection of lakes in the US based on a database of cyanobacteria identification compiled by the NRDC (Natural Resource Defence Council, 2019) were analysed. The database covers events including any type of response that may be related to a harmful algal bloom (HAB) including a cyanotoxin detection or reported illness. With an overall accuracy of 0.88 using German lakes, the validation shows promising results to identify cyanobacteria blooms in inland waters. Further analyses are ongoing to test the algorithm related to high biomass blooms falsely detected as cyanobacteria risks and the cyanobacteria identification within rivers.
The strength of the presented approach is in the implementation of a comprehensive trained random forest model, independent from atmospheric correction uncertainties, covering various cyanobacteria conditions in inland waters. However, some challenges are still present in the model approach, such as uncertainties for high biomass bloom without cyanobacteria or the missing information about the cyanobacteria concentration. While the approach is under development, if proven robust, the setup would allow for an expansion towards a risk assessment of cyanobacteria in inland waters for the high-resolution sensor Sentinel 2 MSI.

Spatiotemporally consistent data on forest height support carbon accounting, forest resource management, and monitoring of ecosystem functions at continental to global scales. Forest conversion, degradation, and restoration monitoring in the context of REDD+, the Paris Agreement, and the Glasgow Declaration requires annual forest height data time series. Forest height has traditionally been directly measured in the field or using airborne laser scanning (ALS). Lately, a sample of near-global forest height data was collected by the NASA Global Ecosystem Dynamics Investigation (GEDI) spaceborne lidar operating onboard the International Space Station (ISS). Recent advantages of Landsat data archive processing and application of the machine-learning tools enabled operational annual forest height monitoring through the integration of the Landsat observation time series with GEDI and ALS-based forest height measurements.
The Landsat data archive is the only tool that enables global multidecadal forest monitoring at medium (30 m) spatial resolution. Our team processed the entire global archive of the Landsat data from 1997 to 2020 to an analysis-ready dataset (GLAD ARD) that supports global and annual forest structure modeling. The GLAD ARD product represents a 16-day time series of normalized clear-sky surface reflectance and brightness temperature. We transformed the annual Landsat reflectance time series into a set of multitemporal metrics (reflectance distribution and phenology statistics) that support spatiotemporal consistency of the forest height modeling.
The Landsat optical data time series do not allow direct measurements of forest height. Instead, the forest structure variables are modeled by relating reference lidar-based forest height measurements to multitemporal spectral data. Non-parametric machine learning tools, such as regression tree ensembles, enable empirical model implementation. The geographic extent of the model calibration is one of the parameters that control model accuracy; locally calibrated models have higher accuracy compared to continental or global models. A near-global sample of forest height collected by the GEDI instrument supports the calibration of local models to create a single global product. We mapped global forest height by calibrating a separate regression tree ensemble model for each GLAD ARD 1x1 geographic degree tile using training data collected from the target and neighboring tiles. We generated 11,860 individual ensemble models within the GEDI data range; the neighboring tiles were defined using a 12-degree radius. Implementation of the overlapping locally trained models ensured spatial consistency of the output product and high accuracy of the map.
To select the GEDI percentile of waveform energy relative to the ground elevation (RH) metric as training data, we compared the RH metrics with the ALS-based forest height metrics at 30 m spatial resolution. The comparison was done for the regions where both ALS and GEDI data were available. For the global forest height model, we used the 90th percentile of the ALS-based canopy height as the reference. The GEDI RH95 metric had the highest correlation with the ALS-based forest height and was selected for the global forest height modeling. The ALS-based 90th percentile height, however, performed poorly in the temperate and northern managed forests, where tall trees are often left within the clearcut area. Using the 75th percentile ensured the correct representation of forest loss within the logging sites. The best matching GEDI metric for the 75th percentile ALS height is RH85, and this metric was used as calibration data to model forest height in Europe.
The GEDI data extent is limited by the ISS orbit, and no data are available north of the 52nd parallel. The ALS data are not universally available, prohibiting local model calibration in boreal forests. We employed two approaches for forest heigh modeling in northern forests. For the global map, we implemented the regional models to predict forest height in the North. Three continental models (North America, Europe, and Asia) were calibrated separately using GEDI RH90 data between 52nd and 40 parallels and manually added non-forest training over tundra and wetlands. For the forest height mapping in Europe, where the ALS data were available at national and sub-national extent for Norway, Sweden, Finland, and Estonia, we have implemented the locally trained models using both ASL 75th forest heigh percentile and GEDI RH85 metrics as calibration data.
Both global and European continental models were implemented to map forest height change from the year 2000 to 2020. To map the global forest height change, we applied the model calibrated for the year 2019 to selected years and calculated the forest height value for the years 2000 and 2020 as the median of 2000, 2001, 2003, and 2017, 2019, 2020 products, respectively. We implemented extensive filtering of the year 2000 and 2020 maps to reduce errors in the model outputs due to remaining atmospheric contamination and differences in the radiometric resolution of Landsat sensors. For the European continental model, we extracted calibration spectral reflectance data from multiple years over stable forests to ensure multitemporal model consistency. The continental model was applied annually to create a pan-European 21-years forest height time series.
The year 2019 global forest height map was compared to the GEDI set-aside validation data. The comparison yielded RMSE of 6.6 m and MAE of 4.45 m, confirming model suitability for forest monitoring and carbon accounting applications. We expect that the year 2000 and 2020 products that are using the same regression model and similar Landsat metrics have similar model uncertainties.
We also validated the forest extent and change using a statistical sample of reference data collected through visual interpretation of the Landsat and high spatial resolution data time series. Forest extent was defined using a 5 m forest height threshold. For the global time-series product validation, we used a reference sample of 1,000 units (individual Landsat 30-m pixels). The sample was allocated using a stratified random design. The strata represent stable forest, non-forest land cover, forest loss, gain, and the possible change omission. The result shows the high accuracy of the year 2000 and 2020 products, with user’s and producer’s accuracies above 94%. The forest loss and gain accuracies are lower, which illustrates the uncertainty of forest change detection, specifically, within open canopy dry seasonal forests. The accuracy of the year 2020 product for Europe was estimated using a stratified sample of 600 pixels and showed the high quality of the continental dataset; user’s and producer’s accuracies are above 87%. The forest change product validation in Europe is ongoing.
The global forest height maps for years 2000 and 2020 confirm the net forest extent loss reported by the FAO Forest Resource Assessment 2020. Both the global map (using the 5 m height threshold to define forest extent) and the FAO data show that the forest area declined by 2.4% during the past 20 years. The FAO forest extent estimate is 0.9% higher compared to our map-based estimate for both years 2000 and 2020. The national-scale year 2020 forest area estimates are also comparable between FAO and our map, yielding r^2 of 0.98 for countries with at least 10,000 ha of forests. Spatiotemporal forest height data allows us to quantify not only the total forest extent change but also the change of the extent of tall, high carbon forests that are frequently converted to shorter tree plantations or secondary forests in the tropics. Our global data shows that the area of tall forests (defined using 20 m height threshold) declined by 4.1%, nearly twice as fast as the total forest extent decline. Our findings, methods, and global and continental products provide tools supporting the implementation of international agreements toward sustainable forest use and climate change mitigation.

Radar satellite imagery from the Sentinel-1 missions is routinely used to map new disturbances in the primary humid tropical forest at 10 m spatial scale and in near real-time (weekly updates). Sentinel-1’s cloud-penetrating radar provides gap-free observations for the tropics, enabling the rapid detection of small-scale forest disturbances, such as subsistence agriculture and selective logging across large forest regions. The RADD (Radar for Detecting Deforestation) alerts were developed in cooperation with Google and the World Resources Institute Global Forest Watch program. The RADD alerts are currently operational for 45 countries across South America, Africa and Insular Southeast Asia, and are available openly via http://www.globalforestwatch.com and http://radd-alert.wur.nl. Accuracy assessment in the Congo Basin yielded high performance of the method with an estimated user’s and producer’s accuracy of ≥ 95% for events > 0.2 ha (Reiche et al., 2021).
We will provide an overview of new developments of the RADD alerts and lessons learned from expanding the RADD alerts to South America. Method advancements will be presented that include, e.g., Sentinel-1 pre-processing workflows (Mullissa et al., 2021), improving disturbance detection using radar texture, and expanding the RADD alerts to other forest ecosystems (incl. dry tropical forests). We will also provide an overview of research progress beyond the timely detection of the location of new forest disturbances into the characterization of different disturbance types. This includes the combination of multiple operational alerts, characterization of drivers, and the rapid monitoring of local carbon losses when combined with locally calibrated biomass estimates (Csillik et al., in review). We will further reflect on the use of the alerts in specific applications with a particular focus on tracking selective logging in the Congo Basin.


Reiche, J., Mullissa, A., Slagter, B., Gou, Y., Tsendbazar, N-E., Odongo-Braun, C., Vollrath, A., Weisse, M. J., Stolle, F., Pickens, A., Donchyts, G., Clinton, N., Gorelick, N., Herold, M. (2021) Forest disturbance alerts for the Congo Basin using Sentinel-1. Environmental Research Letters 16, 2, 024005. https://doi.org/10.1088/1748-9326/abd0a8.
Mullissa, A., Vollrath, A., Odongo-Braun, C., Slagter, B., Balling, J., Gou, Y., Gorelick, N., Reiche, J. (2021) Sentinel-1 SAR Backscatter Analysis Ready Data Preparation in Google Earth Engine. Remote Sensing 13, 10, 1954; https://doi.org/10.3390/rs13101954.
Araza, A. B.; Castillo, G. B.; Buduan, E. D.; H., Lars; Herold, M.; Reiche, J.; Gou, Y.; Villaluz, M. Gabriela Q.; Razal, R. A. (2021): Intra-Annual Identification of Local Deforestation Hotspots in the Philippines Using Earth Observation Products. Forests 2021, 12, 1008. https://doi.org/ 10.3390/f12081008.
Csillik, O., Reiche, J., De Sy, V., Araza, A., Herold, M. (under review) Rapid monitoring of local carbon losses in Africa’s rainforests.

Accurate information is needed to characterize tropical moist forest cover changes, to support conservation policies and to quantify their contribution to global carbon fluxes more effectively. Global and pan-tropical maps have been derived from Landsat time series to quantify global tree cover losses or tropical moist forest (TMF) changes (Hansen et al. 2013, Kim et al. 2014, Vancutsem et al. 2021). However, short-duration degradation and small disturbances (less than 0.09-ha size) are not depicted due to the limiting spatial resolution and frequency of observations of Landsat imagery. Sentinel 2 data bring finer spatial resolution and higher temporal frequency that can support more accurate forest monitoring, in particular for capturing degradation events.

To map the extent and changes of the TMF cover at 10 meter spatial resolution, we developed an expert system that exploits the multispectral and multitemporal attributes of the Sentinel 2 imagery in combination with historical information provided by Landsat to identify deforestation and degradation events since year 2015 (when S2 observations are available).
The mapping method includes four main steps: (i) single-date multispectral classification of Sentinel 2 and Landsat (7 and 8) scenes into four classes (potential moist forest cover, potential disruption, water and invalid observation), (ii) analysis of trajectory of changes from 1990 to 2021 combining the temporal information of S2 and Landsat and production of a “transition” map, (iii) identification of subclasses (tree plantations and mangroves) based on ancillary information and visual interpretation, and (iv) production of a change map at 10m resolution from year 2020 to year 2021.

For the single-date classification, multispectral clusters are defined first by establishing a spectral library capturing the distribution of spectral signatures of a few mainland cover types (moist and dry forests, savanna, bare soils, urban areas, irrigated and non-irrigated cropland, flooded vegetation, snow, ice and water) and atmosphere perturbations (clouds, haze, and cloud shadows) over the pantropical belt. An initial set of 28000 pixels was selected and labeled through visual interpretation of Sentinel 2 data to represent these land cover and cloud classes and an additional complementary set of 25000 pixels was extracted automatically using the TMF map of year 2020.

In the second step of the mapping approach, the temporal sequence of single-date classifications is analyzed for each pixel to determine the extent of the TMF domain at 10m resolution for the year 2020 and then to identify the change trajectories over the period 1990-2021 with six main transition classes: (i) undisturbed forest, (ii) degraded forest, (iii) deforested land, (iv) forest regrowth, (v) other land cover.

The uncertainties of area estimates from this new S2-TMF map will be produced from an independent reference sample of 6000 plots that has been created through the interpretation of most recent high resolution image that is available from Google Earth platform and Planet imagery. The accuracy of the S2-TMF map will be compared to the accuracy of the Landsat-TMF disturbance map that is 91.4% (Vancutsem et al. 2021).


References
● M. C. Hansen, P. V. Potapov, R. Moore, M. Hancher, S. A. Turubanova, A. Tyukavina,
D. Thau, S. V. Stehman, S. J. Goetz, T. R. Loveland, A. Kommareddy, A. Egorov, L. Chini,
C. O. Justice, J. R. G. Townshend, High-resolution global maps of 21st-century forest cover
change. Science 342, 850–853 (2013).
● D.-H. Kim, J. O. Sexton, P. Noojipady, C. Huang, A. Anand, S. Channan, M. Feng, J. R. Townshend, Global, Landsat-based forest-cover change from 1990 to 2000. Remote Sens. Environ. 155, 178–193 (2014).
● C. Vancutsem, F. Achard, J.-F. Pekel, G. Vieilledent, S. Carboni, D. Simonetti, J. Gallego, L. E. O. C. Aragão, R. Nasi, Long-term (1990–2019) monitoring of forest cover changes in the humid tropics. Sci. Adv. 7 (2021), DOI: 10.1126/sciadv.abe1603

Tropical deforestation is the main driver of the biodiversity crisis, contributes to climate change, and results in the widespread degradation of ecosystem services that are critically important for local communities. A major cause of deforestation in the tropics is the expansion of various forms of agriculture, including smallholder agriculture, agroforestry, or agribusinesses cropping and ranching. These diverse forms of agriculture are themselves embedded in a range of social-ecological and institutional contexts. Together, this produces complex deforestation patterns, with major variation in the severity, speed, and spatial patterns of forest loss. Navigating and structuring this complexity remains a major challenge for sustainability science, and hinges on robust, satellite-based data describing how deforestation takes place and how deforestation frontiers advance.
We developed a novel methodology, based on time series of annual forest extent, forest loss, and post-deforestation land cover, to derive a new generation of satellite-based metrics to describe deforestation frontier processes. We showcase this approach for the world’s tropical dry forests, which harbor some of the most rampant deforestation frontiers, yet remain understudied and neglected by policy-making and conservation planning. Specifically, we derive a set of frontier metrics to characterize deforestation frontiers for (1) the entire South American Gran Chaco, based on the entire Landsat archives for 1985 to 2020, and (2) for all tropical dry forests globally, based on Global Forest Watch data from 2000 to 2020. Using time series analyses on stacks of annual forest maps, we derive frontier metrics capturing different aspects of advancing deforestation frontiers, including baseline forest, fractional forest loss, the speed of deforestation, forest fragmentation, the level of activeness of deforestation, current extent of forest, and the post-deforestation land-cover trajectory. Together, these metrics allow to identify key types of frontiers and to map these at high spatial resolution.
For the Gran Chaco, our results show that more than 19.3 million ha of forests were converted to agriculture between 1985 and 2020. Our frontier metrics revealed a heterogeneous pattern of slowly advancing frontiers, often associated with the expansion of ranching, and rampant frontiers (i.e., frontiers of high speed and severity), often associated with the expansion of cropping. Over much of the time period we studied, pansion for ranching dominated, yet cropland expansion into forests was a major deforestation process during the mid-2000s in the Gran Chaco. Importantly, we found widespread areas initially cleared for ranching that were eventually converted to cropping, highlighting the need for considering post-deforestation land uses for better linking frontier dynamics to the underlying processes and for attributing deforestation to commodities.
Expanding our approach to the global scale confirmed that the Gran Chaco contains some of the most rampant deforestation frontiers globally. However, several other areas, especially in South America and Asia, are also characterized by rampant deforestation frontiers. Clearly, such frontiers are associated with expanding commodity agriculture, highlighting the potential of supply-chain policy response steer deforestation, and the need to provide robust monitoring data for that purpose. Additionally, our analyses revealed many deforestation frontiers that are currently in their early stages, particularly in Africa, translating into an urgent need for forward-looking sustainability planning as frontier dynamics unfold. Finally, despite high diversity in frontier dynamics, our approach identified five high-level frontier archetypes that occur globally, paving the way for comparative research and for cross-regional learning.
Satellites map land cover and all satellite-based indicators therefore require some level of translations to infer knowledge on land use change. Our concept of frontier metrics, based on high-resolution forest-cover indicators, provide a robust, repeatable and transferable way for how this translation process can be implemented to move towards a deeper process-based understanding of land-use change. Our approach can identify high-level, recurring frontier types and can therefore be a step towards more context-specific monitoring and policy responses to deforestation.

Tropical forests compose only one third of the global forest estate, but are critical to maintaining biodiversity, abating climate change, and sustaining human livelihoods. Yet, tropical forests are among the most rapidly vanishing biomes on the planet. From the dry forests of the South American Chaco to the tropical moist lowland forests of South-East Asia, tropical forests experience intense pressure from deforestation and forest degradation, and are thus at the center of global conservation initiatives. Yet, tropical moist and dry forests are important in different ways, making their reliable distinction crucial. For example, whereas tropical moist forests harbor disproportionately large portions of species and living carbon stores, tropical dry forests sustain the livelihoods of hundreds of millions people worldwide, regulating freshwater water supplies and delivering other ecosystem services in the world’s drylands. Despite their important, but distinct roles for global sustainability and their exposure to extreme threats, there is a surprising scarcity of data enabling a reliable distinction of these biomes, or a reliable quantification of their respective long-term change rates. Global assessments largely rely on overlays between generic forest-change maps and coarse biome masks. Tropical dry forests dynamics have a particularly poor documentation, with strong disagreements between available maps of tropical dry-forest extents, and no reliable data on long-term global changes.
We will present the first global wall-to-wall data product distinguishing annual spatial extents and changes of tropical moist and dry forests at 300-m resolution back into the early 1990s – a period of adamant importance for tropical forest monitoring, as it marks peaks in both tropical and dry forest losses in several regions worldwide. We developed this product by hindcasting Hansen et al.’s Global Forest Change (GFC) time-series using a multi-scale, spatially and temporally explicit machine learning approach and fusing multiple remote sensing products derived with different satellite sensors. We built our predictive model with more than 600,000 samples, collected systematically in every country to reflect regional differences in forest-management practices and long-term change patterns and drivers. We performed a spatially explicit cross-validation, including a comparison with yearly, high-resolution forest cover data mapped by an independent dataset. Moreover, we collated over 500,000 field plots distinguishing moist and dry tropical forests to infer local boundaries between the two forest types. Our historical forest mapping approach achieved high performance accuracies (R=0.90 RMSE=9.78%), while our forest classification approach performed similarly well in distinguishing tropical moist forest (F1-score=0.91) and tropical dry forest (F1-score=0.92).

The EU Forest Strategy for 2030 calls for an EU-wide integrated forest monitoring framework using, among others, remote sensing technologies. As part of this, the Commission’s EU Observatory on deforestation, forest degradation, changes in the world’s forest cover, and associated drivers is developing Earth-Observation-based monitoring tools for forests. Here, we highlight recent advances from the Observatory, particularly regarding the monitoring of European forest disturbances using remote sensing. Various components of a broader observatory system are namely being developed and tested on pilot case studies before eventual deployment for the entire EU territory. We provide a technical overview of these components, illustrate the latest research developments and show new forest-related prototype products. Reporting forest dynamics on an annual basis requires accurate mapping, characterization and causal attribution of forest disturbance events. To this end, we developed spatio-temporal methods that exploit the full multi-dimensional nature and potential of Copernicus data, while maintaining scalability for use over large areas. These novel approaches rely on the combination of a supervised deep learning model for accurate detection of disturbance events and radiative transfer modeling for detailed disturbance characterization. The large volumes of highly specific training data required for such approach are generated following a semi-automatic iterative protocol. The model performances and accuracy resulting maps are then quantified using a new multi-purpose reference set that will be made publicly available, along with the training data. Radiative transfer models are used to quantify changes in biochemical and structural plant traits associated with forest disturbances. Another key component of the system will be its ability to continuously monitor and provide near real time alerts related to forest anomalies. To efficiently compare a set of existing near real time change detection algorithms (CCDC, EWMA, CuSum/MoSum (also known as BFASTmonitor) and IQR) in an unbiased way, we implemented them all in a single package optimized for fast computation and with a common standardized interface. Results of this comparative assessment which allow for informed decisions on how to best monitor forests for different EU regions are also presented. In the future, we expect the spectrum of this system to further widen around the core components presented here, acting as a catalyzer for research and development on forest monitoring and management by a variety of stakeholders.

INPE intends to describe the highlights of the accomplishments of the Copernicus cooperation arrangement.

This presentation will give an overview of the ongoing cooperation between NASA/USGS, ESA and European Commission on Optical Land Monitoring.

The authors will present current Copernicus capacity building and cooperation initiative based on Copernicus cooperation arrangements in Asia, Africa and Latin America.

The Digital Twin of the Earth (DTE) will help to advance the monitoring, forecasting and visualisation of natural and human activity globally. High-resolution models will track the health of the planet based on a wide range of data, perform simulations of Earth’s interconnected system with human behaviour and thus support the field of sustainable development, reinforcing our efforts to create a better environment.
As one of ESA’s DTE Precursors, our project has supported ESA in defining the wider DTE concept by establishing the scientific and technical basis to realise elements of a DTE in the food systems vertical. The project, run by CGI, and in close collaboration with Oxford University Innovation, IIASA and Trillium, has focused on developing a Food Systems Digital Twin by linking end to end models from forcing meteorology though crop modelling to price impact assessment. This allows for testing policy options linking climate impacts, food production and sustainability, for example by assessing impacts and potential supply vulnerabilities from large scale crop re-zoning designed to enhance biodiversity. Our use case has involved prominent use of AI processing, challenges of model integration at different scales and ingestion of socio-economic as well as physical measurements, thus testing a number of DTE concepts. The end-to-end chain provides decision support outputs with innovation at each stage and have been tested in consultation with potential stakeholders.

The purpose of our use case has been to demonstrate the value of the DTE concept to the scientific community, by integrating the outputs of novel algorithms. We used a machine learning based extreme precipitation model to feed a Global Gridded Crop Model, and after regional downscaling integrated the result into cropland land use and pricing models. The potential benefits of these links include improvements in routine monitoring with regular seasonal assessments, contributions to short term policy responses to crop shortages due to extremes and support to long term policy development to apply appropriate incentives for land rezoning. Architecture and integration considerations for the DTE within the demonstration help to define the next development priorities as part of the roadmap.

The Digital Twin of the Earth as a whole aims to encompass the full information supply chain from data acquisition and data management, though data fusion and information extraction into decision support. This is the scope for ESA and the European Commission aim to reach with the use of Digital Twins, and the impact is aiming to not only be far reaching, but also to be used as a trusted source for sustainability-focused decision making.

In our Climate Impact Explorer we have built a prototype Digital Twin Earth system that brings together advanced Land Surface Modelling (JULES) of African soil moisture, processed using High Performance Computing infrastructure (on JASMIN - https://jasmin.ac.uk) and optimised via data assimilation (LAVENDAR) using state-of-the-art Earth Observation data (soil moisture and solar-induced fluorescence). We have developed a Machine Learning emulator on top of this system to enable fast exploration of a wider climate space, driven by ISIMIP-based climate scenarios, without costly model simulations. We condense these complex soil moisture outputs into key drought metrics relevant to our stakeholders and make them available via an interactive data portal and Jupyter Notebook environment hosted on JASMIN’s cloud. This system enables decision makers without expert technical knowledge to generate and visualise decision relevant drought information relating to regionalised impacts of climate change.

This work has been carried out as part of the ESA Digital Twin Earth Precursors activity. With the short timescale for the project (12 months) and emphasis on innovation and pioneering new applications it was essential to have the required computing resources easily at our disposal. The combination of an established HPC environment alongside cloud computing resources and large storage capacity on JASMIN meant that it was possible to commence development activities from the outset. This was assisted in large part by the Cluster-as-a-Service system available on JASMIN’s Cloud. This provides a web user interface to rapidly deploy a shrink-wrapped environment from pre-prepared templates - in this case templates for the deployment of Pangeo (Jupyter Notebook service with Dask) and an Identity Service to manage and authenticate users to the platform.

Data from the HPC system was output as regular netCDF files one per time step onto traditional POSIX storage. Using object storage it was possible to make outputs readily available to the cloud environment in analysis-ready form. This entailed serialisation of the data in Zarr format and rechunking to suit time series-based data queries, the predominant access pattern for analysis. This was critical in enabling the creation of responsive interactive map-based web user interfaces.

The ICT provisioned via JASMIN facilitated effectively an incubator environment for the Digital Twin demonstrator and in this respect mirrors the essential elements identified in the DestinE high-level architectural blueprint of an open core platform providing HPC, data sources and cloud computing capability.

The Forest Digital Twin Earth (Forest DTE) will be a data-driven, physical-coherent, approach to Earth system science, which will make use of existing Earth Observation (EO) capabilities and physically-based modelling, to create a digital replica of the world’s forests.
A precursor of the system was created by a consortium funded by the European Space Agency following the Destination Earth (DestinE) initiative. The consortium included VTT Technical Research Centre of Finland, Department of Forest Sciences of the University of Helsinki, Simosol OY, Unique GmbH, Cloudferro Sp z o.o., and Romanian National Institute of Forest Research (INCDS). DestinE, a part of the European Green Deal, aims to develop a high precision digital of the Earth to monitor and simulate natural phenomena and related human activities. Forest DTE, a specialized digital twin, would provide detailed information on the functioning, climate effects and carbon exchanges related to the forests, which cover approximately one-third of the planet's land surface. Satellite-based estimates of forest structure and above-ground biomass offer the only means to obtain homogeneous and extensive information on the state of the world's forests. Other information sources, such as field plot data, soil maps, climate predictions and forest management scenarios are needed to understand the ongoing biological and anthropogenic processes, and to predict the state of the forests in the future.

Before the precursor implementation, we identified the needs of forest sector users on the specific data products, and spatial and temporal resolutions of the Forest DTE. In 2020, the most important forest variables were related to the carbon stocks of the forests, the necessary temporal time scales were from seasonal or yearly changes to several forest successions (i.e., several hundreds of years), and the required spatial resolutions generally corresponded that of high-resolution optical imagery, with a need of spatial aggregations to forest management or administrative units. The Forest DTE Precursor was implemented on the Forestry Thematic Exploitation platform, hosted on the CREODIAS cloud, for selected and rather limited test areas in Europe. However, the existing computational facilities were found to be sufficient also for large-scale implementation of a forest digital twin at the spatial and temporal scales required by the user.

Based on the experience obtained during the precursor implementation, the following limitations related to user needs need to be addressed when implementing the full Forest DTE: 1) Availability of homogeneous (i.e. containing the minimum set of required variables and confirming to the basic quality standards) forestry field data in a computer-readable format; 2) Validation capabilities of the full Forest DTE chain at the spatial resolution of the products; 3) Improved determination of species or plant functional type and their proportions in mixed stands; 4) Integration with other components of the Digital Twin of the Earth: the relevant technical tools and protocols need to be developed. These forest-related limitations need to be overcome within a decade to reach the goal of DestinE of having a functional Digital Twin of the Earth in ten years.

Although field-measured forest data are scarce for many regions of the world, national forest inventories exist in many countries and can provide data, although sometimes at irregular intervals. In the forthcoming decades, robust machine learning-based tools need to be created to make best use of these data, match the field measurements with satellite observations, and allow reliable estimation of key forest variables from Earth observation data. The next step in Forest DTE requires a modeling of forest functioning at the spatial scale of very high resolution satellite sensors. In principle, tools for this exist, but contrary to forest variable estimation, direct validation of the forest productivity models at this spatial resolution, and at the temporal scales required by the users, is still a scientific challenge.

The validation calls for the use of different data sets of carbon and other fluxes, computed at vary different spatial and temporal and spatial resolutions, e.g. from atmospheric model inversion. A validation of a forest DTE, or any DTE in general, implies a full unification of top-down and bottom-up approaches: the fluxes computed from detailed measurements and simulations of the terrestrial and aquatic ecosystems need to match the global atmospheric simulations. New data and methods may need to be included for this. For example, satellite-borne measurement chlorophyll fluorescence will allow to create a global map of photosynthesis. Future hyperspectral constellations will allow a more detailed mapping of overstory species and, potentially, plant stress. In order to make a functional digital twin of the Earth, all these very diverse data sources will need to be integrated on a single platform.

When finally functioning as a part of the Digital Twin of the Earth, Forest DTE will act as a spatially explicit simulation tool, which can be initialized using a snapshot of data, including EO imagery, environmental data, and field measurements of forestry parameters. It will give users tailored access to high-quality information, services, models, scenarios, forecasts and visualizations as required by DestinE.