Spatial Synchronization and Propagation of Soil Moisture Droughts Over China Using Complex Network Theory

Abstract Investigating the spatial association and evolution of soil moisture (SM) droughts is crucial for mitigating their negative impacts on agriculture, the ecosystem, and the economy. Using complex network theory and random forest models, this study analyzed the spatial synchronization characteristics, regional drivers, and propagation patterns of weekly SM drought events across China from 1979 to 2019. Results revealed drought events displayed higher spatial synchronization in eastern Tibet, northern Shaanxi, and southeastern and northeastern China, with average degree centrality (DC) exceeding 250. The drivers of drought event synchronization exhibited regional heterogeneity: precipitation (P, 24.1%), potential evapotranspiration (PET, 15.2%), and atmospheric pressure (PRS, 10.8%) were dominant in western China; P, zonal wind speed at 10 m height (WIN_10v), and terrain gradient contributed more to DC simulation in the northeastern region, with an average contribution of 7.7%; river area (RiA) was the primary factor affecting drought synchrony in southern China, contributing 15.1%, which was 2.4–3 times higher than relative humidity (RHU) and WIN_10v. Drought propagation analysis identified source regions in western Xinjiang and northern Tibet while sink zones were concentrated in eastern Tibet, western Sichuan, northern Guizhou, northeastern Jiangxi, and southeastern Heilongjiang Provinces. Droughts originating from northern Tibet predominantly propagated along the south‐southeast (SSE) direction, with a strength of 90 and distances ranging from 1,500 to 1,800 km. This study is valuable to enhance the understanding of the spatiotemporal patterns of SM droughts in China and highlight the potential of complex network approaches for agricultural drought monitoring and early warning.