Considering different water supplies can improve the accuracyof the WOFOST crop model and remote sensing assimilation in predicting wheat yield

The study was carried out in order to clarify the effects of different water and irrigation conditions on crop models and remote sensing assimilation results. It involved taking winter wheat from 17 test sites in Henan Province as the research object and calibrating the World Food Studies model. The ensemble Kalman filter algorithm was used to calibrate the two modes and Moderate-resolution Imaging Spectroradiometer-Leaf Area Index of the calibrated world food studies model. The study found that the average error of the world food studies model for simulating flowering and maturity periods is within 2 days, the R2 of the leaf area index calibration results is between 0.87-0.98, and the R2 and root mean square error of the verification results are 0.77 and 1.06 respectively. Under the latent model, the R2 of the world food studies model taking account of the water supply situation and the assimilation results without taking account of the water supply situation are 0.50 and 0.48, respectively. In the water restric-tion mode, the R2 increased from 0.79 to 0.86 compared with the assimilation results where the water supply was not considered. The results show that: depending on the water supply of different regions, selecting the corresponding assimilation parameters can effectively improve the prediction accuracy of crop models and remote sensing assimilation for wheat yields under different water and irrigation conditions.