Propagation characteristics of soil moisture drought to hydrological drought by spatial zoning and dynamic Bayesian network

Understanding how soil moisture drought transitions into hydrological drought is essential for assessing drought risk, where existing studies on drought propagation often overlook regional heterogeneity and dynamic dependencies among drought variables. Thus, this study proposes an integrated framework to characterize propagation characteristics from soil moisture to hydrological drought. The k-means algorithm was used to conduct the spatial zoning to construct drought indices SSMI and SRI. The sliding window-based Spearman’s correlation analysis on these two indices was applied to estimate the drought propagation time. Dynamic Bayesian Network (DBN) and Copula functions were combined to model their dynamic relationships and infer the drought propagation probability. Results indicate that, during 1982-2023, drought propagation from soil moisture to hydrological droughts in the Yangtze River Basin is highly non-linear with substantial regional heterogeneity. The average drought propagation time is 7.39 months, with pronounced lag effects. Drought propagation probability also revealed distribution differences and distinct hierarchical response patterns, with averaging 19.13% in high-elevation upper reaches compared to 10.95% in middle-lower reaches, and 16.52% under extreme soil moisture droughts compared to 12.33% under light droughts. Comparative evaluations demonstrate that spatial zoning based SSMI and SRI can more accurately reflect regional drought dynamics, and DBN can also enhance the sensitivity to drought variables compared to BN. The proposed probabilistic reasoning framework can help to strengthen scientific decision-makings in agricultural production and ecological conservation.