REFERENCE EVAPOTRANSPIRATION ESTIMATION USING ADAPTIVE NEURO-FUZZY INFERENCE SYSTEMS

Abstract

Evapotranspiration, an important component in terrestrial water balance and net primary productivity models, is difficult to measure and estimate. In this Study, the potential of the adaptive neuro-fuzzy inference system (ANFIS) is investigated in modelling of daily grass crop reference evapotranspiration (ETo) obtained using the Penman-Monteith equation. Various combinations of daily climatic data, namely solar radiation, air temperature, relative humidity and wind speed, are used as inputs to the ANFIS so as to evaluate the degree of effect of each of these variables on daily Penman-Monteith estimated ETo. The results of the ANFIS model are compared with a multiple linear regression model. Mean square error, average absolute relative error and determination coefficient statistics are used as comparison criteria for evaluation of the model performance. The ANFIS technique whose inputs are solar radiation, air temperature, relative humidity and wind speed, gave mean square errors of 0.016, average absolute relative errors of 6.4%, and determination coefficients of 0.996 for Morgan Hill 139 station (San Francisco Bay, USA). Based on the comparisons, it was found that the ANFIS model could be successfully employed in estimating the daily ETo. Copyright (C) 2008 John Wiley & Sons, Ltd.