WaSCIA Demonstration for the African Stakeholders
Evapotranspiration and soil moisture computation
All the development are made on the Web Advanced Space Developer Interface (WASDI) platform, which is a cloud platform for Earth Observation (EO) that helps both remote sensing experts and end users work with geospatial data. Dedicated processing nodes have been provided with the European Space Agency (ESA) Network of Resources (NOR); and are functioning correctly. This processing capacity will be available until the end of the project in October 2024.
The model developed for the measurement of the soil moisture and evapotranspiration was presented by RSS-Hydro. Automated algorithms have been developed to derive these two characteristics. However, a compromise should be made between the size of the processed area and the computation time. The primary proposition is to run the models over 300km² areas and to decide of the best update frequency for training the Random Forest model for downscaling (bi-monthly or monthly). Different validation tests have already started, based on MET data (of ESA) for the evapotranspiration, or on AGRHYMET data for the soil moisture. Overall the results are satisfying but will be improved to reach validation thresholds.
The next steps are the implementation of the algorithms on WASDI and the optimization of the different algorithms, in terms of methodology and computation time.
Climate indices development
The work is based on the combination of past observations with numerical weather models to generate times series of multiple climate variables. A list of different climate indices has been selected based on the results of the User Survey. The development of the code to compute the climate indices is ongoing. Data preparation and quality control have also been carried out. The next steps are also to deploy the algorithms on WASDI.
Evapotranspiration (left) and Soil moisture (right) plots for Senegal.
Available soil moisture measuring stations in Senegal.
Discussions and feedback
The demonstration received positive feedback from LPAOSF concerning the methodology that has been developed and the results presented. The evapotranspiration and soil moisture are indicators that are really pertinent for Senegal. Such data are barely available over large areas, while they could be really valuable in a wide range of domains; in particular for the management of water resources and the development of sustainable agriculture.
DGPRE also emphasized the quality of the work performed and the progress made. The robustness of the presented methodology was appreciated. DGPRE asked how the results could be extrapolated to take account of the spatial and climatic differences all across Senegal. He also wondered how the methods would be validated in the absence of in-situ data. On this last aspect, RSS-Hydro plans to use other sources of data to validate the results (e.g. Sentinel-3 products) for situations where no in-situ is available. However, for a few samples or locally, they will take benefit of the availability of in-situ data.
Finally, AGHRYMET also gave their acknowledgment for the work done. They raised the possibility to get in contact with other entities working in West Africa, in order to get data for validation. The Regional Center is also wondering about how capacity building will be carried out, on the use of WASDI but also on the algorithms and programming languages. On this last point, the team explained that most of the project activities planned for 2024 will be dedicated to training and to enable knowledge transfer; the project workshop is also planned to be held in Senegal in 2024.
AGHRYMET would also need to deploy the platform on its own cloud environment to enable direct access for neighboring countries. This will be under further discussion; those perspectives, that will allow the platform to be used beyond the project, will be discussed at ESA premises, with ESA representatives during the next project meeting to be held together with the users in December 2023.