Anfis Based Reference Evapotranspiration (ET0) Estimation Using Limited and Different Climate Parameters

This study aims to estimate the reference evapotranspiration (ET0) with adaptive neuro-fuzzy inference system (ANFIS). The calculation of ET0 requires climate data such as maximum and minimum temperature, wind speed, sunshine duration and humidity. Sometimes it may not be possible to access the climate data. In the study, ET0 is estimated with ANFIS by using fewer input parameters. Moreover, it is proved that it is possible to calculate ET0 via ANFIS by using some other climate parameters such as temperature-humidity-wind index (THW), air pressure, wind chill, which have no direct effect on the calculation of ET0. In the study, five scenarios were created and evaluated with a statistical performance indicator. In Scenario 1, ET0 was estimated using THW, air pressure and air density. The relationship between the ET0 estimated by ANFIS and the calculated ET0 was found to be 0.76 (R2). In Scenario 2, THW, solar energy and solar radiation were used and R2 was found to be 0.66. Scenario 3 used THW and wind chill, resulting in an R2 of 0.43. In Scenario 4, THW and maximum temperature were used and R2 was 0.87. In Scenario 5, THW and humidity were used and the R2 with ET0 was 0.84.