Application of the Thornthwaite Water Balance Model to Evaluate the Accuracy of Actual Evapotranspiration Data of GLEAM and GLDAS Databases in the Western Sub-basins of Lake Urmia

Actual evapotranspiration (AET) plays an important role in agricultural water management, drought forecasting, and climate change projection. In recent years, the use of indirect methods to estimate AET has led to the development of various databases. The existence of various uncertainties in these data makes their evaluation necessary. In this study, focusing on the historical period of climate studies (1976-2005), the estimated values of AET by means of the Thornthwaite water balance model were used as a basis for evaluating GLEAM and GLDAS databases, in Nazlou Chay, Rozeh Chay, Shahr Chay, and Barandouz Chay sub-basins in the west of Lake Urmia. The results showed that on an annual scale, the difference between GLEAM and GLDAS databases was not noticeable (RMSEGLEAM=0.13-0.46 mm/day, RMSEGLDAS=0.13-0.48 mm/day). But on a seasonal scale, the performance of GLEAM and GLDAS data differed in different seasons. Based on the evaluation criteria, the GLEAM database performed better than the GLDAS (GLEAM: RMSEWinetr=0.11-0.33, RMSESpring=0.25-1.15, RMSESummer=0.58-0.77, RMSEAutumn=0.15-0.33; GLDAS: RMSEWinetr=0.15-0.27, RMSESpring=0.32-1.14, RMSESummer=0.64-0.95, RMSEAutumn=0.15-0.40). Also, in annual and seasonal scales, GLEAM and GLDAS data were more reliable in Shahr Chay sub-basin (RMSEShahrchay=0.12-0.58, RMSENazlouchay=0.15-0.85, RMSERozehchay=0.19-1.15, RMSEBarandouzchay=0.11-0.77). According to the results of the present study, it is not recommended to use GLEAM and GLDAS data in hydrometeorological studies in Nazlou Chay and Rozeh Chay sub-basins. The results of the present study can be used in the studies of projecting AET future changes, as a basis for evaluating the outputs of climate models as well as correcting their bias.