Evaluation of SEBS and SEBAL algorithms for estimating wheat evapotranspiration (case study: central areas of Khuzestan province)

Abstract

AbstractThis study aimed to accurately estimate daily wheat evapotranspiration using two remote sensing algorithms, Surface Energy Balance System (SEBS) and Surface Energy Balance Algorithm for Land (SEBAL), in central Khuzestan province during 2019–2020. The results of two algorithms were compared with lysimeter (as a direct method), FAO-Penman–Monteith (FAO-PM), two temperature-based methods (Hargreaves-Samani and Blaney-Criddle), two radiation-based methods (Priestley–Taylor and Doorenbos–Pruitt), and two mass transfer-based methods (Mahringer and World Meteorology Organization) (as indirect methods). Coefficient of Determination (R2), Root-Mean-Square Error (RMSE), Percentage of Bias (PBIAS), Mean Bias Error, Mean Absolute Percentage Error, and Nash–Sutcliffe indicators used for comparing the results. According to the results, both SEBAL and SEBS algorithms showed the highest compatibility with lysimeter data (R2 = 0.92 and 0.96, RMSE = 2.15 and 1.53 mm/day, respectively). Comparing both algorithms with the FAO-PM method, resulted in RMSE and R2 of 2.42 mm/day and 0.87 for SEBS and 3.14 mm/day and 0.79 for SEBAL. The Hargreaves-Samani method (R2 = 0.72, RMSE = 16.4 mm/day) and (R2 = 0.8, RMSE = 10.4 mm/day) among temperature-based methods, Doorenbos–Pruitt (R2 = 0.71, RMSE = 3.33 mm/day) and (R2 = 0.79, RMSE = 2.63 mm/day) among radiation-based methods, and the Mahringer method (R2 = 0.6, RMSE = 6.8 mm/day mm/day) and (R2 = 0.68, RMSE = 5.51 mm/day) among mass transfer-based methods yielded better estimations than SEBAL and SEBS algorithms, respectively. Owing to the high accuracy of SEBAL and SEBS algorithms, in estimating the amount of evapotranspiration in the study area and close to the actual values in the field, using energy balance algorithms is recommended in Khuzestan province.