Automatic rice cropping intensity mapping in cloudy and foggy areas leveraging simplified harmonic analysis based on sentinel-1 imagery

Accurate mapping rice cropping intensity and cropping area is crucial for global food security, water resource managing, as well as economic stability. However, in cloudy and foggy rice-producing areas, current automated mapping of rice cropping intensity is significantly restricted by a lack of sufficient training samples, diverse backscatter patterns of rice, low efficiency in time-series feature extraction, and similar phenological evolution of rice and non-rice crops. To tackle the above issues, this study constructed a depolarization index (DI) time series that can stably track rice phenology, integrated periodicity assessment and cosine component weighting to simplified harmonic analysis method, and introducing DI increment assessment and temperature constraints to eliminate interference from non-rice crops. The rice cropping intensity and cropping area in southern China and the five Southeast Asian countries were mapped at a resolution of 10 m on the Google Earth Engine (GEE) platform. The overall accuracy (OA) was 84.07%, which was highly consistent with the provincial agricultural statistics (R2>0.79), performed significantly improved mapping accuracy and enhanced detail capture, while also revealed significant regional differences in the effectiveness of temperature constraints across climatic zones. The proposed framework enriches the algorithm library of rice cropping intensity mapping that does not rely on training samples, providing valuable methodological insights for large-scale rice monitoring in complex environments. The latest release of 10 m rice cropping intensity map is publicly available at https://doi.org/10.5281/zenodo.15095946.