Compatibility of Hybrid Neuro-Fuzzy Model to Predict Reference Evapotranspiration in Distinct Climate Stations

The aim of this study is to model Reference Evapotranspiration (ET0) in Nigeria and Cyprus with Maiduguri and Larnaca as a case study region. Adaptive Neuro Fuzzy Inference System (ANFIS) which utilized 3 membership function owing to its fine mapping capability was employed for the modeling purpose. Multiple Linear Regression (MLR) model was also developed. The results were compared to Penman-Monteith (FAO-56-PM) model. Monthly average of long-term climate data including minimum temperature, maximum temperature, relative humidity, and wind speed were used as inputs to the models. The performance of the models was evaluated by two global statistics of Root Mean Square Error (RMSE), and Determination Coefficient (DC). The results indicated that ANFIS had better performance than MLR models. The results also showed ANFIS was capable of modeling ET0 in the study regions efficiently, but had better performance in Maiduguri than in Larnaca region.