Prediction Model for Daily Reference Crop Evapotranspiration Based on Hybrid Algorithm in Semi-Arid Regions of China

The accurate estimation of reference crop evapotranspiration (ETO) plays an important role in guiding regional water resource management and content research. In order to improve the accuracy ETO prediction regions with missing data, this study used partial correlation analysis method select factors that have a large impact on as input combinations construct models for typical stations semi-arid China. A biological heuristic optimization algorithm (Golden Eagle (GEO) Sparrow (SSA)) Extreme Learning Machine model (ELM) were combined accuracy. results showed Ra was primary factor affecting model, importance range 0.187–0.566. Compared independent ELM hybrid has higher stability. estimated value SSA-ELM under five-factor condition (Ra, RH, Tmax, Tmin, U2) is closest standard calculated by FAO56 PM: RMSE = 0.067–0.085, R2 0.998–0.999, MAE 0.050–0.066 NSE 0.998–0.999. general, combination can be estimate high reducing factors. Use first five extracted from build based China’s recommended, which also provide similar regions.