Knowing the water status of a crop can lead to sustainable water use. With an increase in droughts and competition for water, crop water status and need will become more relevant to water managers in the future. Canopy temperature (Tc) is a good indicator for water status due to its inverse relation to the rate of canopy water loss but is sensitive to changing weather conditions. To take them into account, the Crop Water Stress Index (CWSI) was developed a long time ago and has been shown useful for water stress monitoring in semi-arid regions. Different approaches to calculate CWSI have been developed based on differences in how the upper and lower limits are derived, but it is presently unclear which method works best for assessing crop water status of potato in humid regions and how robust the relationships are with in situ measurements of crop water availability. The objectives of this study were to: i) Derive and assess the performance of three CWSI methods (CWSIe - empirical, CWSIt - theoretical, CWSIh - hybrid) as a proxy for crop water stress through relations with volumetric soil moisture (SM), ii) Assess the transferability of derived CWSI-θ relations from one year to another year and its applicability to drone-based TIR images for mapping soil moisture at field scale, iii) Evaluate the functioning of CWSI as a water stress indicator in a humid region on potato, a crop susceptible to water stress.
An experiment was carried out in Hamerstorf in Northern Germany (annual rain totals of 622 mm, mean annual air temperature of 8.8°C, silty sand is the prime soil type). Different irrigation water treatments on potato crops were carried out over the two years 2018 and 2019: Optimum irrigation (OP hereafter), aiming to highest yield by allowing no drought stress for the plants, Reduced irrigation (RD hereafter), aiming to reduce water consumption with a minimum of yield loss, and no irrigation as control treatment. Six thermal infrared thermometers (Thünen Institute in-house development) were set up at six plots to record continuous canopy temperatures (Tc) of three RD and three OP treatments from June to September for each year. For each treatment three plots were monitored, and Tc recorded every 3 seconds and averaged over 30 minutes. The mean values of the three OP or RD treatments were used for stress analysis. Continuous SM measurements were made at RD and OP positions covering the top 60 cm soil profile in intervals of 10 cm. Meteorological measurements of air temperature (Ta), solar irradiance (Rs), relative humidity (H), and wind speed (WS), were measured every two minutes and aggregated to thirty minutes intervals. Daily rainfall was recorded at a weather station located on the experimental site. Additional meteorological variables such as vapor pressure deficit (VPD) and net radiation (Rn) were calculated using standard FAO method. Visible/near infrared (VNIR) and Thermal Infrared (TIR) images were collected over the experimental plots on several days during the growing season and pre-processed according to standard procedures. The CWSIe was calculated according to Idso et al. (1981) using the single day method and the multi-day method (Gardner et al. 1992). CWSIt was calculated according to Jackson et al. 1981. For calculating CWSIh, the empirical upper limit was substituted into the equation for calculating the lower limit, with the consequence that measurements of net radiation, ground heat flux and aerodynamic resistance were no longer required. R2 between midday (11:00 – 15:00 h) CWSI and SM at 10 cm depth were reported.
Results showed that CWSI works only when high radiative heating is accompanied with increasing air dryness (Rs > 600 Wm-2 and VPD > 20hPa). This confirms that CWSI works in a humid region, but under situations similar to the semi-arid conditions which it was originally developed for. The predictive performance of different stress metrics in relation to SM increased in R2 from Tc (R2=0.58) to Tc-Ta (R2=0.76) to CWSI (R2=0.74-0.87) (situation in 2019). CWSI-SM relations calibrated in one year (2019) and applied to another year (2018), and vice versa, revealed absolute errors of 1-3% SM; such accuracy is considered good enough to support irrigation management. The three investigated types of CWSI, i.e., CWSIe, CWSIt and CWSIh performed similarly, but have different input requirements; CWSIh could be a promising alternative to the traditional CWSI as it combines the aspects of CWSIe (empirical upper limit) and of CWSIt (theoretical lower limit) which has certain advantages. TIR observations should ideally be acquired in the noon to afternoon hours between 12 and 16 hrs daytime, which confirms previous findings from other studies. Finally, the drone-based water stress images and the SM maps (derived from the developed CWSI-SM relation) captured well the applied irrigation patterns and could help to decide when to irrigate and how much water to apply.

Accurate and spatially explicit information on irrigation is essential for sustainable water resource management, crop condition monitoring, and modelling. Although some advances have been made for irrigation mapping using remotely sensed data, most studies are conducted in semi-arid areas, while field-level irrigation mapping remains challenging for temperate regions. However, the importance of information about irrigation in temperate regions is increasing, as irrigation requirements are expected to increase over the coming years․ To assess the applicability of different time series for irrigation mapping at field scale, we used optical and Sentinel-1 (S1) data over northern Germany, an area characterized with heterogeneous field sizes, crop patterns, irrigation systems, and management. An extensive amount of field-scale irrigation data reported by farmers was collected and used as a reference for model training and validation. We used the Framework for Operational Radiometric Correction for Environmental monitoring (FORCE) to produce spectral-temporal metrics from integrated Landsat and Sentinel-2 time series. The derived feature set comprises several vegetation indices and Tasseled Cap components which were aggregated over the growing season and specific key phenological stages of winter and summer crop growth. Further, MODIS-based evapotranspiration was integrated to account for water use. Irrigated areas were then classified using the temporal aggregates from remotely sensed time series using random forest (RF) and gradient boosting (XGboost) -based classifiers. In general, XGboost outperformed RF for irrigation mapping with both composites from specific growth stages and growing seasons. Overall accuracy reached satisfactory levels (~74%). Besides the cross-validation of the classification results, the area of applicability of spatial prediction models was estimated. Irrigation probability for maize and potato was generally higher than for cereals, for which supplementary water supply does not necessarily result in an economically viable yield increase. The combined use of optical and radar data enhanced the classification accuracy. Overall, the study presents a framework for field-level irrigation mapping over large areas, which can inform crop models and assist the estimation of water resource demands. Further research will be focused on the transfer of the methods to other areas.

The agricultural sector is the biggest and least efficient water user, accounting for around 80% of total water use in South Europe, which will be further impacted by climate change in the incoming years. Precision agriculture tools are then needed to increase water use efficiency.
Here, the proposed system couples together remotely sensed land surface temperature (LST), leaf area index (LAI) and ground soil moisture data (SM) with a pixel wise crop-water-energy balances model, for improving irrigation management. The SAFY (Simple Algorithm for Yield) crop model has been fully coupled with the energy water balance FEST-EWB model, exchanging in a double direction the LAI evolution in time from SAFY, which is used by FEST-EWB for evapotranspiration computation, while FEST-EWB provides soil moisture (SM) and LST to SAFY model for constraining crop growth.
A data assimilation framework, based on the Ensemble Kalman filter approach, is implemented to reduce the requirements for parameters calibration, either for soil assimilating satellite LST and for crop growth using LAI. This framework allows overcoming the issues related to crop exposure to shocks due extreme events non-reproducible by the model alone, as well as nutrient lack, crops hybrids or precise amount of irrigation water.
The FEST-EWB-SAFY model has been applied in two Irrigation Consortia in the North and South of Italy which differ for climate and agricultural practices, using data from Sentinel2, Landsat 7 and 8 satellites. The model has then been validated in specific fields where ground measurements of evapotranspiration, soil moisture and crop yields are available.
Overall, the results suggested that the under-calibrated model estimates of LST, LAI, SM and yield are enhanced through the assimilation of satellite data, suggesting the potential for improving irrigation management at both field and Irrigation Consortium scales.

Policy makers around the world are confronted with increasingly complex decisions to sustain the future of agriculture systems to meet the growing demand for food along with challenges caused by increasing water scarcity and impacts of the climate crisis. Monitoring of water productivity in agriculture, that is the amount of crop which can be produced by using a certain amount of water, is necessary to reduce the externalities of the agricultural sector while ensuring food and water security for all.
FAO and partners maintain a publicly accessible near real time database developed with the analysis of high-resolution satellite images in conjunction with specific algorithms to determine spatial and temporal variability of agricultural water and land productivity. The database called WaPOR provides near real time information through an open access data portal (WaPOR https://wapor.apps.fao.org) enabling range of service providers to assist farmers to attain more reliable yields and improve their livelihoods. Irrigation operators have access to new information to assess the performance of systems and to identify where to focus investments to modernize the irrigation schemes, and government agencies can use the information to monitor and promote the efficient use of natural resources.
To translate the data into actionable and policy relevant information, remote sensing, hydrology and agronomy experts are developing targeted case studies in selected countries in the framework of the project Water Productivity Improvements in Practice (WaterPIP). In Egypt, for example, data from WaPOR is being used to compare the water productivity of two sugar crops, sugar beet and sugar cane, starting from bio-physical parameters but also considering socio-economic implications.
The session will describe the ways in which the FAO WaPOR project, now in its second phase, is supporting national institutions in monitoring and improving water productivity. It will then focus more in detail on the analysis done on water productivity of sugar crops in Egypt. The implications of temporal and spatial resolution of input RS products for monitoring WP will be discussed, particularly in relation to the estimation of actual evapotranspiration and the need for surface temperature (LST) products at appropriate spatio-temporal scales.

Observing and understanding the climate change and human impacts on rice agriculture in the Mekong delta to support mitigation and adaptation measures

Thuy Le Toan1, Alexandre Bouvet 1, Stephane Mermoz 2, Hoa Phan 1, Thu Trang Le 1, Hironori Arai 1, Thierry Koleck 3, Lam Dao Nguyen 4, Nguyen Quoc Khanh 5
1. Centre d’Etudes Spatiales de la Biosphère, Toulouse, France
2. GlobEO, Toulouse, France
3. CNES, Toulouse, France
4. Vietnam Space Center, Ho Chi Minh city, Vietnam
5. Vietnam Remote Sensing Department, Hanoi, Vietnam


Abstract

The countries in the Mekong region and in particular the deltas in the region are among the world most exposed to the threats of climate change, which are being worsened by stressors from fast growing population. In the last decades, the frequency and intensity of inundations, droughts, salinity intrusion have increased, causing negative impacts on different sectors, the most important being rice agriculture, which plays a central role in the social and economic development of the countries in the region. This is the case of the Vietnam Mekong Delta.

In order to support stakeholders to define adequate adaptation and mitigation strategies, it is essential to develop methods and to conduct studies, to monitor the impacts of climate change and human pressure on rice agriculture, in order to advance our understanding on the causes of the changes observed on rice land.
In most research work, the focus has been mainly on the impacts of climate change on crop productivity. However, for crop production, the impact could be more important on the crop harvest area which undergoes significant changes in the last few years, resulting from reduction of the number of crops per year, conversion of rice crop into other land use (aquaculture) or other perennial or annual crops. The drivers of these changes could be either the impacts of climate change on the habitat suitability for rice crop, or/in conjonction with the socio economic conditions in the region, both could be at the origin of farmers’s adaptation measures.

With the unprecedented observations collected by Sentinel-1 satellites since 2015, the role of EO data is reinforced in the monitoring of the full agricultural dynamics at national scale down to the single fields. In this study, the dynamic information on rice area estimates, rice growth stage and agricultural practices, as well as on the extent of seasonal flood on the Mekong Delta, are produced and the changes observed on rice land in the last 6 years are derived.
Inter-annual changes in land use, changes in rice harvest area, change in cropping density, and changes in crop calendar have been quantified and mapped. The most significant changes are the losses of dry season rice harvest area observed in 2016 and 2020. Other changes included rice fields converted into aquaculture, turning from triple rice crop to double crop, or having shifted calendar -up to 2 months from the traditional calendar. These changes appeared in scattered fields or small parts of administrative units, and are apparently decided by farmers as autonomous adaptation to climate change effects.
The changes have been analysed, in combination with climatic and environmental data, to improve our understanding on the drivers of these autonomous adaptation.
To understand the drivers of these changes, linked either to climate or socio-economic factors, the changes are analysed against a range of data including climatic data, sea level rise, flood extent, drought index , salinity intrusion…. The analysis has pointed out that most observed changes can be explained primarily by climatic factors. For example the loss of rice harvest area in coastal region in spring 2016 and spring 2020, following El Niño events, has been found in regions having high drought index and high saline water concentration. In regions impacted by upstream flooding, the flood extent and duration are larger and longer than before during the flood season, changes are observed in the calendar of the crops before and after the flood season. In the regions impacted by drought, salinity intrusion or flood, a number of rice fields were converted into shrimp farms, or let fallow during the critical season.
In order to support the stakeholders for longer term mitigation and adaptation measures, we estimate the suitability for rice cultivation in the future, based on the projections of climatic factors under different climatic scenarios for 2030 and 2050 developed in the Gemmes Vietnam project (Espagne et al., 2021).
In particular, the projected flood maps have been established, based on the projected elevation map of the Mekong Delta (Minderhoud et al., 2019), and taking into account the cumulative effect of Sea Level change and land subsidence, the latter being caused mainly by groundwater extraction. The results show that without adaptation, a large part of the rice land will fall below sea level. For example 34% of rice area in the Mekong Delta will be lost by submersion in 2050, for a Sea level rise of 25 cm. The first mitigation measure would be to halt groundwater extraction rate at the impacted provinces. The adaptation measure would be building hard infrastructure to protect the land from inundation, but in this case, there is a need to consider socio-economic and ecological factors.
The projection of saline water intrusion (Eslami et al., 2021) has been used to map the projected rice land which will become less suitable for rice cultivation in 2030 and 2050, The impacted area, with the water salinity exceeding 2‰, amounts 10.5% of the dry season rice area. The mitigation measure would be to reduce sediment starvation from upstream hydropower dams and excessive sand mining, which increase saline water intrusion. The adaptation measure would be the use of salt resistant species, or conversion of rice cultivation into other land use, e.g. aquaculture.
Finally, the ambition of our studies is to provide tools to support stakeholders in Vietnam for devising adaptation and mitigation measures, to simulate on the effects of each of the relevant driving parameters. Further works need to be done integrating socio-economic factors.
Acknowledgment
The study benefits from the results of the works conducted in the ESA GEORICE project, the AFD GEMMES project, and the CNES VIETSCO project.
References

ESA-GEORICE : https://www.globeo.net/georice

Eslami, S., Hoekstra, P., Minderhoud, P. S. J. ., Trung, N. N., Hoch, J. M., H.Sutanudjaja, E., Dung, D. D.,
TranQuang, T., Voepel, H. E. , Woillez, M.-N. and van der Vegt, M. (2021). Projections of salt intrusion in a mega-delta under climatic and anthropogenic stressors. Nature Communications Earth & Environment, 2(1), 1-11. doi:10.1038/s43247-021-00208-5

Espagne E. (ed.), T. Ngo-Duc, M-H. Nguyen, E. Pannier, M-N. Woillez, A. Drogoul,T. P. L. Huynh, T. Le Toan , T. H. Nguyen, T. T. Nguyen, T. A. Nguyen, F. Thomas,C. Q. Truong, Q. T. Vo, C. T. Vu. 2021. Climate change in Viet Nam- Impacts and adaptation. A COP26 assessment report of the GEMMES Viet Nam project. Paris. Agence Française de Développement, 1 November 2021.
https://www.afd.fr/en/ressources/gemmes-vietnam-climate-change-impacts-and-adaptation

Minderhoud, P. S. J., Coumou, L., Erkens, G., Middelkoop, H., & Stouthamer, E. (2019). Mekong delta much lower than previously assumed in sea level rise impact assessments. Nature communications, 10(1), 1-13.

VietSCO/Vimesco: https://www.spaceclimateobservatory.org/vimesco-rice

Mountain grasslands in the European Alps play a crucial role in climate regulation, biodiversity safeguarding, landscape conservation, and soil quality preservation. The ownership of managed grassland is usually highly fragmented, and the management practices are very heterogeneous. Despite water being abundant in the past, water scarcity has started to raise concerns in the Alps because droughts are becoming more and more frequent, causing unstable income for mountain farmers. Risk management instruments, like insurances, are necessary to sustain grassland-based mountain agricultural systems and to allow them to overcome production shortcomings, and thus maintain their functioning over time. Traditional insurance schemes would require yield measurements and physical inspections by insurance appraisers to assess damages, but this approach is not economically sustainable due to their high cost compared to the low value of the production. Index-based insurance can overcome these issues because payoffs depend on an index related to grassland production that does not require physical checks. In this context, high-resolution satellite data from the Sentinel-2 constellation allow the development of accurate and low-cost tools to support risk management. In the project DRI2 (DRought Insurance - phase 2), we estimate yield losses due to drought in mountain grasslands by Sentinel-2 satellite data. DRI2 will be the basis for a digital insurance scheme that the agricultural consortia will test for the growing season 2022 over the Provinces of Bolzano and Trento in north-eastern Italy. The project builds on a close dialogue between researchers and stakeholders, including public administrations and agricultural consortia, which helps to match the needs of the forage production sector in the region of interest. We exploit a combination of Leaf Area Index (LAI) and a meteorological water stress coefficient as a proxy for grassland biomass production. We estimate LAI from Sentinel-2 satellite data by the SNAP biophysical processor and we investigate different gap-filling methods under cloudy sky conditions. We calculate the grassland production index as the growing season cumulate of the daily product between LAI and water stress. Finally, we estimate the drought index as the anomaly of the production index based on the preceding five years. We aggregate the index at the farm level based on the digital cadastral data available from the local authorities. To verify the ability of the model to reproduce the actual grassland biophysical parameters, we compare Sentinel-2 LAI with ground-based observations of LAI whose collection plan allows representing the variability of the vegetation within a Sentinel-2 pixel. We also compare Sentinel-2 LAI and measured aboveground biomass to verify whether they follow the same temporal dynamics. The preliminary results show an overall RMSE of 0.8 and r² of 0.83 for LAI, with LAI ranging from 0.5 to 6, based on data collected from the year 2017 to the year 2020 at the Long-Term Ecological Research (LTER) experimental site of Matschertal/Val di Mazia. The comparison with measured wet biomass indicates that Sentinel-2 LAI has a high correlation with biomass, with r² of 0.79, follows the same seasonality, and allows mowing detection. During the next project phase, we will extend the evaluation of the production index to a larger number of test sites, in which data collection is ongoing, to consider different management practices. Furthermore, we will assess the ability of the index to identify yield variations at a few experimental sites where yield data are available for several consecutive years.