In this study, we tried to address the applicability of using dynamic remotely sensed data into a static crop model to capture the yield spatiotemporal variability at the field scale. Taking the example of the crop environment resource synthesis for wheat (CERES-wheat), the model was calibrated, improved, and validated using three years of winter wheat field measurement data (growing seasons of 2017–2019). We assimilated the Landsat-based leaf area index (LAI) into the model using the particle filter approach. Four vegetation indices, including NDVI, SAVI, EVI, and EVI-2, were evaluated to identify winter wheat LAI’s best estimator. A linear regression of Landsat-EVI-2 was found to be the most accurate representation of LAI (LAI = 10.08 × EVI-2 − 0.53) with R² = 0.87, and mean bias error = − 2.04. The higher LAI accuracy from EVI-2 was attributed to the soil and canopy background noise reduction and accounting for certain atmospheric conditions. Assimilating the LAI based on Landsat-EVI-2 into the CERES model improved the model’s overall performance, particularly for grain yield and biomass simulations. The default model predicted LAI_max, grain yield, and biomass at 5.1 cm² cm⁻², 8.3 Mg ha⁻¹, and 14.9 Mg ha⁻¹ with RMSE of 1.44, 0.91 Mg ha⁻¹, and 1.2 Mg ha⁻¹, respectively, while the modified model (using the Landsat-EVI-2 data) predicated these values at 6.6 cm² cm⁻², 9.9 Mg ha⁻¹, and 16.6 Mg ha⁻¹ with RMSE of 0.81, 0.54 Mg ha⁻¹, and 0.62 Mg ha⁻¹, respectively.

在这项研究中，我们试图解决将动态遥感数据用于静态作物模型以捕获田间规模的产量时空变化的适用性。以小麦的作物环境资源综合为例（CERES-小麦），使用三年冬小麦田间测量数据（2017-2019年的生长季节）对模型进行了校准，改进和验证。我们使用粒子滤波方法将基于Landsat的叶面积指数（LAI）同化为模型。对包括NDVI，SAVI，EVI和EVI-2在内的四个植被指数进行了评估，以确定冬小麦LAI的最佳估计量。发现Landsat-EVI-2的线性回归是最准确的LAI表示（LAI = 10.08×EVI-2 − 0.53），R ² = 0.87，平均偏差误差= − 2.04。来自EVI-2的更高的LAI精度归因于土壤和冠层背景噪声的减少以及考虑到某些大气条件。将基于Landsat-EVI-2的LAI纳入CERES模型可改善模型的整体性能，尤其是在谷物产量和生物量模拟方面。默认模型预测LAI_最大值，谷粒产量，和生物质在5.1厘米² 厘米^-2，8.3镁公顷^-1，和14.9镁公顷^-1 1.44 RMSE，0.91镁公顷^-1和1.2镁公顷^-1，修改后的模型（使用Landsat-EVI-2数据）分别将这些值设为6.6 cm ²  cm ^-2。，9.9镁公顷^-1，和16.6镁公顷^-1为0.81 RMSE，0.54镁公顷^-1，和0.62镁公顷^-1，分别。