Estimation of daily reference evapotranspiration using artificial intelligence methods and empirical equations

Evapotranspiration is an important parameter in hydrological and meteorological studies. Accurate estimation of evapotranspiration is crucial in planning and operating water structures. In this study, the daily reference evapotranspiration amount (〖ET〗_0) was calculated using the Penman-Monteith equation accepted as the standard equation by the FAO (Food and Agriculture Organization). Daily reference evapotranspiration was tried to be estimated using Adaptive Network Based Fuzzy Inference System (ANFIS), M5 Decision Tree (M5T), Multiple Linear Regression (MLR) methods, Turc traditional equations and the results were compared with each other. Station (USGS No: 330716096592400) near Lake Lewisville (Texas, USA) was chosen as the study area. Average daily air temperature (T), wind speed (U), solar radiation (SR) and relative humidity (RH) were used for daily average evapotranspiration estimation. The coefficient of determination (R2), root mean square error (RMSE) and mean absolute error (MAE) were used to evaluate the model results. Model results indicate that ANFIS, M5T and MLR methods were successful in estimating the daily evapotranspiration amount of the region.