Evaluation of evapotranspiration estimations of GLEAM model in northern part of Karkhe basin.

Regional estimation of evapotranspiration (ET) is crucial in agricultural water management, cropping pattern and crop yield forecasting.The main purpose of this study is evaluation of actual evapotranspiration estimations of Global Land Evaporation Amsterdam (GLEAM) model of Karkhe basin, southwest of Iran. GLEAM outputs include different set of RS-based estimations, i.e. transpiration, interception, bare soil evaporation, open water evaporation and snow sublimation. In current research, the GLEAM estimation of actual evapotranspiration were compared by water balance approach (assuming a watertight basin) using coefficient of determination (R2). The corresponding values of R2 for the each study stations across the Karkhe basin namely Aran-e-Gharb,Pirsalman,Nazarabad,Polchehr,Paypol,Poldokhtar,Tangsazbon, Pole-e-kaskhan were 0.66, 0.62,0.77,0.84,0.59,0.49,0.56 and 0.66 Similarly, the RMSE values were 0.7, 0.62, 0.74, 0.25 0.45, 0.87, 0.73 and 0.51. The MAE of same comparisons were found to be 0.52, 0.43, 0.58, 0.18 0.3, 0.71, 0.53 and 0.36. Besides the relative deviation of GLEAM dataset estimations and those obtained from water balance method were calculated as -21,-3.73,-4.96,-4.19,-19.61,-2.39,-3.64,-25.18. The best agreement by all statistical indices was observed in Polchehr sub-basin. The results of this study indicated that RS-based estimations of GLEAM model can be used in Karkhe basin, especially in regions with lack of sufficient observed data.