Improving the Satellite-Based Winter Wheat Monitoring by Using UAV-Derived Leaf Chlorophyll Content

Mapping regional leaf chlorophyll content (LCC) still faces challenges of the scale mismatch between ground sampling points and satellite pixels. To improve the satellite-based model, this study explored the potential of using high-resolution unmanned aerial vehicle (UAV) data to bridge the scale mismatch. We collected LCC measurements, UAV, and Sentinel-2 (S2) data on 2 March 2024. First, both linear and random forest models developed from UAV and S2 data at 0.1 and 10 m, respectively. Second, the UAV-derived LCC predictions were used as reference data with S2 VIs to build S2-UAV-based models at 10 m using different amount of training data. Finally, the contributions of UAV data were investigated in UAV and satellite synergy for mapping regional LCC in winter wheat. These results showed that the S2-UAV-based models significantly outperformed the models directly developed by S2 and ground data (R² = 0.34–0.61), achieving higher accuracy (R² = 0.57–0.80). Due to spatial heterogeneity, ground measurements can hardly represent the mixed information in an S2 pixel and underestimate model performance of predicting LCC. By contrast, the UAV data with high spatial resolution and high accuracy of LCC predictions can bridge the scale mismatch between ground sampling points and S2 pixels by expanding training data and improving spatial details. Overall, this study demonstrated that UAV and satellite synergy provided proper solutions for timely and accurate crop monitoring at regional scale.