Better Performance of the Modified CERES-Wheat Model in Simulating Evapotranspiration and Wheat Growth under Water Stress Conditions

Crop models are important for understanding and regulating agroecosystems. Although the CERES-Wheat model is an important tool for winter wheat research, it has some limitations under water stress conditions. To narrow the gap, this study aimed to improve the performance of the CERES-Wheat model under water stress in arid and semi-arid regions based on the winter wheat experimental data from 2012 to 2014. The Priestley–Taylor (PT) and FAO56 Penman–Monteith (PM) equations were used to calculate the reference crop evapotranspiration and further modified the crop coefficient of the CERES wheat model to improve the simulation accuracy of crop yield and evapotranspiration under water stress conditions. The results showed that: water stress before jointing seriously affected the accuracy of the CERES-Wheat model in simulating biomass and grain yield, so it was necessary to improve the original model. In the original and improved models, the accuracy of the PM equation was lower than that of PT. In addition, the simulation accuracy of the improved model was higher than that of the original model (the average RMAE and RRMSE are less than 30%). In general, among the four scenarios, the PT equation for calculating crop reference evapotranspiration and crop coefficient had the best performance. Water stress occurred at the heading and grain filling stages, and the simulated biomass was in good agreement with the observed results, which better simulated the soil water content under water stress at the later growth stages. Therefore, the change in water stress response function had positive effects on winter wheat growth under simulated water stress conditions. This study provided a reference for applying the CERES-Wheat model in arid and semi-arid areas.