Evaluation of a generalized linear model for the actual evapotranspiration using satellite and reanalysis data

Abstract

An important issue for agricultural planning is to estimate evapotranspiration accurately due to its fundamental role in sustainable use of water resources. It is essential to have reliable and precise evapotranspiration (ET) measurements to improve models or products. This work aims to evaluate a generalized linear model (GLM) in order to estimate actual evapotranspiration of barley crop with satellite (Landsat, Sentinel, and CERES) and reanalysis (MERRA-2) data. The results obtained were compared with water balance values from an agrometeorological station. The GLM with the combination of MERRA-2/CERES/Sentinel 2 as input was the best performance (R2 = 0.59). The results show the feasibility of applying machine learning algorithms for obtaining actual evapotranspiration values in agricultural plains without ground agro-meteorological data.