Field Crop 2025, Vol.8, No.6, 293-300 http://cropscipublisher.com/index.php/fc 297 application can all be carried out in a targeted manner. Sometimes, just by taking a glance at the anomalies in the spectral time series data, one can identify the problem points in advance. Although it may seem complicated, for farmers, this means more precise decision-making, higher output and more resource conservation. Figure 2 Flowchart for the winter wheat yield estimation using the EnKF-based assimilation algorithm (Adopted from Zhuo et al., 2019) 6.2 Improved accuracy over isolated remote sensing or modeling alone If relying solely on remote sensing, although the images are clear, it is difficult to explain the physiological processes behind them. Relying solely on models is also easily restricted by input parameters. After the combination of the two, the improvement in prediction accuracy is almost immediate. For instance, after introducing variables such as leaf area Index (LAI), Normalized Vegetation Index (NDVI), or enhanced Vegetation Index (EVI) into the model, R² increased, RMSE decreased, and the estimation accuracy of spatial yield variance even improved by 70%, with the error reduced by more than half (Li et al., 2024). With the assistance of artificial intelligence algorithms such as CNN-GRU and ensemble learning, the system remains stable when facing complex and rapidly changing environments (Wang et al., 2023). It can be said that this is a way to make data and models "complement each other's strengths and weaknesses". 6.3 Applications in precision agriculture, food security, and policy-making The potential of this integrated system is far more than what looks good in the laboratory. It can draw field yield maps and help farmers adjust management plans according to plot differences (Yang et al., 2024); It can also issue early warnings before the signs of reduced crop yields emerge, providing a basis for grain reserves and market regulation (Zhang et al., 2024). For government departments, it can also generate high-precision spatial data for formulating agricultural plans, risk assessments or insurance policies. In other words, from farmland to the policy level, this approach makes it truly possible to "accurately identify and effectively manage". 7 Challenges and Limitations 7.1 Data quality issues: cloud cover, satellite revisit times, input data gaps No matter how advanced remote sensing is, it cannot escape The Times when the weather is not cooperating. When thick clouds cover the area, the images captured by satellites become incomplete, and data from key growth periods are often missing or distorted as a result. What's more troublesome is that the revisit cycle of satellites does not always happen to coincide with the important growth nodes of crops. Sometimes, missing a few days can
RkJQdWJsaXNoZXIy MjQ4ODYzNA==