Estimating referencecrop evapotranspiration using artificial neural network based on random samples

Quantification of referencecrop evapotranspiration(ET₀) is necessary in the context of many issues, for example calculating crop requirement, scheduling of irrigation and the management of water resource so on. Random samples instead of the long-term climatic data were used as samples to construct Artificial Neural Network(ANN) model for estimating ET₀ and the performance of ANN was compared with the method of modified Penman-Monteith. The maximum and minimum temperature, maximum and minimum relative humidity, solar radiation, wind speed were selected as network inputs and ET₀ was selected as network output. After calculating, the result shows that the estimation of ANN has higher accuracy by error analysis.