High-resolution Mapping of Terrestrial Evapotranspiration using ECOSTRESS: Insights into Surface Energy Balance Modeling

Abstract

ECOsystem Spaceborne Thermal Radiometer Experiment on Space Station (ECOSTRESS) has been providing high spatio-temporal thermal infrared (TIR) observations (~70 m, 1-5 days) since August 2018. Land surface temperature (LST) retrieval obtained from TIR observations indicates the thermal status of the surface as a consequence of the land-atmosphere exchange of energy and water. It carries the imprint of vegetation water use and stress, thus serving as a pivotal lower boundary condition for retrieving evapotranspiration (ET). Taking advantage of the ECOSTRESS observations, the European ECOSTRESS Hub (EEH) funded by the European Space Agency (ESA) retrieves high-resolution ET for terrestrial ecosystems. In EEH Phase 1 (2020-2022), instantaneous ET data between 2018 and 2021 were generated from three models with different structures and parameterization schemes over Europe and Africa, including the Surface Energy Balance System (SEBS) and Two Source Energy Balance (TSEB) parametric models, as well as the analytical Surface Temperature Initiated Closure (STIC) model. The evaluation by comparing against ground measurements at 19 eddy covariance sites for 6 different biomes over Europe showed that the physically based STIC model had relatively better consistency and higher accuracy across varying aridity and diverse biomes. Also, an advantage of STIC was found as compared to the official ECOSTRESS ET product obtained using the PT-JPL model, especially over arid and semiarid regions due to the weak LST control in PT-JPL. Taking advantage of the recalibrated ECOSTRESS Collection 2 data, EEH Phase 2 (2023-2025) analyses the impacts of LST estimates from different algorithms on ET retrieval and related biophysical conductances over different biomes. It is found that ET estimates of STIC driven by LST retrieved from the two most commonly used algorithms (i.e., split window, SW, and temperature and emissivity separation, TES) have comparable accuracies. The sensitivity of ET to LST over savannas is almost three times of those over biomes over lower aridity. Surface-canopy conductance is more sensitive to surface temperature as compared to aerodynamic conductance. Overall, the EEH is promising to provide quality assured ET estimates for monitoring terrestrial ecosystem water use and stress. Furthermore, it will facilitate the preparation for the next generation high-resolution thermal missions by investigating surface energy balance modeling, including TRISHNA (CNES/ISRO), SBG (NASA), and LSTM (ESA).