A climate-informed nonstationary trivariate drought index for regional assessment in Jiangsu Province, China

Study region: Jiangsu Province, eastern China, characterized by a humid monsoon climate, dense river networks, intensive water use, and strong interactions among precipitation, runoff, and groundwater. Study focus: This study develops a climate-informed Nonstationary Trivariate Standardized Drought Index (NTSDI) to characterize compound drought dynamics under changing hydroclimatic conditions. Precipitation, surface runoff, and groundwater are integrated within a unified probabilistic framework. Large-scale climate indices are incorporated as explanatory variables to account for climate-driven nonstationarity. Nonstationary marginal distributions are modeled using Generalized Additive Models for Location, Scale, and Shape, while a time-varying D-vine copula is applied to capture evolving multivariate dependence structures. New hydrological insights for the region: Results reveal a significant hydroclimatic regime shift around 2003 across precipitation, surface runoff, and subsurface runoff. Compared with the pre-change period, precipitation decreased by 8.78%, surface runoff by 20.98%, and subsurface runoff by 36.89%, indicating amplified drought propagation from atmospheric input to subsurface hydrological response. The NTSDI effectively captures compound drought evolution and delayed subsurface hydrological responses, providing a more robust basis for regional drought monitoring than conventional indices. By explicitly accounting for climate-driven nonstationarity and multivariate dependence, it offers valuable support for drought risk assessment and adaptive water resources management in highly dynamic hydroclimatic regions such as Jiangsu Province.