中国10m分辨率小麦种植模式（CPChina）时空分布数据集（2020年）

基于Sentinel-1雷达和Sentinel-2多光谱遥感影像，采用耦合时序特征的作物轮作指数（如VVC-WR、WMCR），建立了中国10m分辨率小麦种植模式（CPChina）时空分布数据集（2020年），实现了全国范围内小麦种植模式的自动化制图。研究显示，2020年中国小麦播种面积达206,919平方公里，其中90%以上采用多熟种植模式，冬小麦–玉米轮作占51.39%，冬小麦–水稻轮作占21.12%，春小麦单作占6.86%。该数据集在无需地面训练数据的情况下，总体制图精度达92.57%，特别适用于多云地区和复杂小农耕作系统。研究成果为精准农业管理提供了高分辨率数据支持，揭示了小麦种植向干旱/半干旱区扩张的趋势（占全国52%），并量化了作物轮作对多熟制农业（占全国45.68%）的贡献，为优化种植结构、提升土壤健康及减少碳排放提供了空间决策依据。该方法体系可为发展中国家构建自动化作物监测系统提供参考。

Based on Sentinel-1 radar and Sentinel-2 multispectral remote sensing imagery, this study developed the 10-meter resolution spatiotemporal distribution dataset of wheat cropping systems in China (CPChina, 2020) using crop rotation indices with coupled temporal features (e.g., VVC-WR, WMCR), achieving automated nationwide mapping of wheat cropping systems. The research shows that China's total wheat sown area reached 206,919 square kilometers in 2020, of which over 90% adopted multiple cropping systems: winter wheat-maize rotation accounted for 51.39%, winter wheat-rice rotation 21.12%, and single spring wheat cropping 6.86%. Without requiring ground truth training data, this dataset achieved an overall mapping accuracy of 92.57%, and is particularly suitable for cloud-prone regions and complex smallholder farming systems. The dataset provides high-resolution data support for precision agricultural management, reveals the expansion trend of wheat planting into arid and semi-arid regions (accounting for 52% of the national total), quantifies the contribution of crop rotation to multiple cropping agriculture (accounting for 45.68% of the national total), and provides spatial decision-making support for optimizing cropping structure, improving soil health and reducing carbon emissions. This methodological framework can serve as a reference for developing automated crop monitoring systems in developing countries.