Using Genetic Algorithms and Gene Expression Programming to Estimate Evapotranspiration with Limited Meteorological Data

In this study, genetic algorithm (GA) was employed to detect the most important variables for estimating ETo among mean temperature (Tmean), maximum temperature (Tmax), minimum temperature (Tmin), sunshine hours (n), relative humidity (RH), and wind speed (WS). The results show that Tmean and WS are the most important meteorological variables to model evapotranspiration in Iran. Then, we selected gene expression programming (GEP) to model ETo based on Tmean and WS historical data. The results indicate that the GEP has good performance for semiarid and Mediterranean climates compared to very humid and some arid regions. In addition, GEP is an effective solution when there is insufficient meteorological data available.