Comparison of Artificial Neural Network and Empirical Models to Determine Daily Reference Evapotranspiration

The accurate estimation of reference crop evapotranspiration (ETo) is essential in irrigation water management to assess the time-dependent status ofcrop water use and irrigation scheduling. The importance of ETo has resulted in many direct and indirect methods to approximate its value and includepan evaporation, meteorological-based estimations, lysimetry, soil moisture depletion, and soil water balance equations. Artificial neural networks (ANNs)have been intensively implemented for process-based hydrologic modeling due to their superior performance using nonlinear modeling, patternrecognition, and classification. This study adapted two well-known ANN algorithms, Backpropagation neural network (BPNN) and Generalizedregression neural network (GRNN), to evaluate their capability to accurately predict ETo using daily meteorological data. All data were obtained fromtwo automated weather stations (Chupungryeong and Jangsu) located in the Yeongdong-gun (2002-2017) and Jangsu-gun (1988-2017), respectively.
Daily ETo was calculated using the Penman-Monteith equation as the benchmark method. These calculated values of ETo and correspondingmeteorological data were separated into training, validation and test datasets. The performance of each ANN algorithm was evaluated against ETocalculated from the benchmark method and multiple linear regression (MLR) model. The overall results showed that the BPNN algorithm performedbest followed by the MLR and GRNN in a statistical sense and this could contribute to provide valuable information to farmers, water managers andpolicy makers for effective agricultural water governance.