Özet:

Evaporation, one of the most important elements of the hydrological cycle. In addition, accurate estimation of evaporation is used as a fundamental parameter in many areas including design of irrigation system, hydrological modeling, river flow estimation and irrigation planning. Although there are different empirical approaches to prediction of evaporation, artificial intelligence methods have also been used. In this study, monthly pan evaporation measured at Adıyaman and Diyarbakır stations in the Southeastern Anatolia Region are estimated using the different climatic parameters such as temperature, wind speed, relative humidity, pressure, solar radiation and monthly open days. Artificial Neural Networks (ANN), Adaptive Network Based Fuzzy Logic Inference System (ANFIS) and Gene Expression Programming (GEP) methods are used to predict pan evaporation with different input combinations. To determine performance of these methods to prediction of pan evaporation, R2, Mean Square Error (MSE) and Mean Absolute Error (MAE) criteria are used. The GEP method is more successful than the other two methods used and the highest R2 values calculated with this method during the test process are 0.923 for Adıyaman station and 0.963 for Diyarbakır station.