A Copula-based composite drought index for vegetation and hydrological droughts

Drought is a complex natural hazard with profound impacts on vegetation and water resources. However, existing monitoring approaches often treat vegetation and hydrological droughts separately, limiting a comprehensive understanding of their interactions. To address this gap, this study developed a novel Copula-based vegetation and hydrological composite drought index (VHCDI) using non-parametric kernel density estimation (KDE) to capture skewed and bimodal distributions in normalized difference vegetation index (NDVI) and the Gravity Recovery and Climate Experiment (GRACE)-derived drought indices. Applied for the first time to the Three-North region (2002–2022), the VHCDI integrates a standardized vegetation index (ZVI) derived from NDVI and a GRACE-based drought severity index (GRACE-DSI) derived from downscaled terrestrial water storage change (TWSC) at 0.25° resolution. The results show that: (1) Despite TWS and precipitation exhibiting decreasing trends in 59.7% and 33.8% of the region, respectively, NDVI increased across 92.2% of the area, largely due to ecological restoration; (2) By capturing nonlinear dependencies, VHCDI combines ZVI’s sensitivity to short-term precipitation anomalies with GRACE-DSI’s ability to reflect accumulated hydrological deficits. Comparative assessments against MCI and scPDSI indicate that VHCDI provides a complementary and effective characterization of drought, capturing both rapid vegetation responses and accumulated hydrological deficits. Its integrative nature offers added value for comprehensive drought monitoring, particularly in regions where vegetation and water storage dynamics are both important. These findings contribute significantly to comprehensive drought monitoring in the Three-North region.