Pan evaporation modelling in Central India using soft computing techniques

Pan evaporation (Epan) is an important parameter in water budget estimations and in modeling crop water response in different weather conditions. It has been widely used as an index of evapotranspiration, lake and reservoir evaporation, potential or reference crop evapotranspiration and irrigation scheduling. Because evaporation is a nonlinear, stochastic and complex process, it is difficult to derive an accurate formula to represent all the physical processes involved (Moghadamnia et al. 2009). In recent years, application of artificial intelligence (AI) techniques, such as multivariate adaptive regression splines (MARS), least square support vector regression (LSSVR) and M5 tree model in estimation of hydrological parameters have been widely considered by the most of researchers. The main objective of the present study is to evaluate the performance of M5 tree, LSSVR and MARS methods in estimating long-term monthly Epan in central India (i.e. Madhya Pradesh state).