Research of artificial neural network model for reference crop evapotranspiration

According to Hetao district long-term meteorology data and reference evapotranspiration (ET₀) which were calculated by Penman—Monteith method, main meteorology data affecting ET₀ were regressed and analyzed. Based on these, four factors input vector (mean temperature, net radiation, relative humidity and wind speed at 2 m high) BP network forecast model about ET₀ were compared with three factors (mean temperature, net radiation, relative humidity) input vector. The research indicated BP network model was suitable for ET₀ forecasting, four-factor and three-factor input vector BP network model were both convenient and feasible for forecasting ET₀ and could satisfy the needs of production. The precision of four factor input vector network model was higher than three factor input network model. This research is the supplement for traditional ET₀ calculation.