Disseminating knowledge in the NENA region on the most widely used Remote Sensing Evapotranspiration models.

The Food and Agriculture Organization of the United Nations (FAO) has started implementing a webinar series on Remote Sensing (RS) determination of evapotranspiration (ET) which pursues four main objectives, including updating the participants’ knowledge regarding current adoptions and the latest advances in RS ET determination methods and approaches, increasing the capacity of key water professionals in the assessment of RS ET uncertainty, improving audiences’ awareness on the strengths and limitations for the range of RS ET models and databases available, and provision of a complete and advanced set of documentation on the various topics treated during the webinar series.

The webinar series have been planned to be presented in four modules. The first module was implemented from March to May 2021, and introduced the most widely used RS ET models. During some of the webinars related to different models, experiences from different countries in the NENA region have been presented, regarding the application of that specific model in a study area. The first module consisted of nine sessions, where the first session provided a background overview on the application of RS for ET modelling, focusing on the potential strength, weakness, pending questions and uncertainties regarding the application of RS in ET mapping. A brief review of concepts of RS ET determinations and basic physics of RS was done, as well as a summary of the satellites and sensors presently used for ET mapping, along with their spectral, spatial and temporal resolutions,  and a vision for future of ET modelling. In the following sessions, different scholars attempted to present the ET RS models developed by their teams, and some experiences in the application of these models. Some general pieces of information regarding these webinars have been provided in the below table. A total of 850 participants attended module 1.

During the webinars of this module, the development of RS ET models included SEBAL, METRIC, ALEXI, SSEBop, ETLook, ETMonitor, ETWatch, SIMS and PT-JPL. The models crossed several approaches, including the surface energy balance (SEB), the Penman-Monteith equation, soil moisture modelling, vegetation indices (VIs) based modelling, on-source and two source (soil/vegetation), etc. RS from satellites (Aqua, Terra, LandSat, Sentinel, etc.) have been utilized in the provision of the different data required for running the models.

Although each of the presented model has its own specific advantages and limitations, spatial and temporal resolutions, strengths and weaknesses, uncertainties and best-fitting domains of application, a few general points can be drawn from the presentations.

-          Satellite data fusion and combination of different resources of satellite imagery could be associated with the production of different parameters, and thereby the ET, with higher spatial and temporal resolutions. These processes can even lead to the combination of different sensors (e.g., multispectral, thermal and radar) and platforms (e.g., satellite, airborne and UAV) at different spatial and temporal resolutions for the implementation of the RS ET models.

-          Despite advances in RS ET determinations, many researchers still emphasize on the use of ground data to calibrate models and validate results.

-          For many of the mentioned models, researchers have developed programs and tools in the open source and cloud computing platforms such as python and Google Earth Engine (GEE) for ET determination or access to ET maps. This includes PySEBAL as a Python module for SEBAL developed by IHE-Delft, Earth Engine Evapotranspiration Flux (EEEFlux) and EEMETRIC as METRIC based tools in GEE for access to ET snapshots, FAO Water Productivity Open-access portal (WaPOR) as a portal to access ET data using ETLook, etc. Accordingly, users with various objectives and levels of knowledge can refer to these tools to implement the RS ET models or receive the required ET outputs.

-          There are still some large sources of uncertainties for the determination of ET using most of the presented models. This can include uncertainty in the calculation of the different parameters, especially LST and air temperature, and presence of the cloudy pixels that cause gaps and biases. Another source of errors in the RS ET models is the impact of time integration from instantaneous to daily, and then to monthly ET maps.

To address the raised issues, it was necessary to conduct further webinars on the introduction of the major databases and portals providing ET derived by some of the models presented in the Module 1, new developments in the models and the applied data, and the RS ET errors and uncertainties, which are presented in the webinars under the modules 2, 3 and 4, respectively. These modules are currently being implemented.

This webinar series is implemented within the framework of the Water Scarcity Initiative of the FAO Regional Office for Near East and North Africa, with the support of the regional project “Implementing the water efficiency/productivity and water sustainability in NENA countries”, funded by the Swedish International Development and Cooperation Agency.

For more information and details on upcoming modules and webinars, please check the webinars page here.