Spatiotemporal analysis of crop cropping patterns in the middle and lower reaches of the Yangtze River using multi-source remote sensing

Abstract

The middle and lower reaches of the Yangtze River are a major grain and oilseed production region in China, where crop planting structure and phenological timing are closely related to agricultural water demand and regional water security. This study developed a multi-source remote sensing framework to map crop types, planting areas, and key phenological stages from 2018 to 2024. Sentinel-1, Sentinel-2, and Landsat-8/9 data were integrated to construct optical, radar, vegetation, water, and yellowness time-series features. Time-weighted dynamic time warping was used to identify rice, wheat, and maize, while a rule-based sample generation and one-class classification method was applied to map rapeseed. Crop phenology was extracted using smoothed vegetation index time series, curvature extrema, and dynamic thresholds. The results show that rice, wheat, maize, and rapeseed dominate the regional cropping system. Rice is mainly concentrated in lake plains and lowland areas, wheat is prominent in Jiangsu and parts of Anhui, rapeseed is widely distributed as a winter rotation crop, and maize is mainly found in Hubei, Hunan, and Anhui. Three planting structure patterns were identified: double-rice dominated areas in the Dongting Lake and Poyang Lake plains, rice-wheat balanced areas in Jiangsu and parts of Anhui, and diversified cropping areas in Hubei. Crop phenology showed clear regional differences controlled by latitude, thermal conditions, cropping systems, and local management practices. The proposed framework provides spatially explicit crop information for agricultural water demand estimation, irrigation scheduling, drought-flood response, and regional water resource management.