A Bayesian Copula Based Integrated Drought Index for Compound Drought Monitoring in India

Drought is a complex and persistent hazard affecting agriculture, ecosystems, economic stability, and public health. Traditional univariate drought indices often overlook the interconnected behavior of drought components, limiting their capacity to support holistic drought assessment and early warning. To address this gap, we develop a Bayesian Copula-Based Integrated Drought Index (IDI) that jointly represents meteorological, hydrological, and agricultural drought conditions across India. The framework integrates a modified Standardized Precipitation Index (SPI), Standardized Runoff Index (SRI), and Standardized Soil Moisture Index (SSMI) at multiple monthly timescales using gridded data at 0.25° resolution from 1951 to 2024. An Archimedean copula family is used to characterize the dependence structure among drought drivers. Marginal distributions are selected based on a rigorous comparison of candidate probability models; Gamma for precipitation and GEV for both streamflow and soil moisture, as validated through the Kolmogorov–Smirnov test and Akaike Information Criteria. Model parameters are estimated through Bayesian inference via the Differential Evolution Markov Chain (DE-MC) algorithm, which combines differential evolution with Markov Chain Monte Carlo sampling to ensure robust, efficient convergence and uncertainty quantification. Comparative analysis demonstrates that the IDI outperforms individual indices in representing the spatial extent, persistence, and severity of drought events. By accounting for multi-source drought information within a probabilistic and dependency aware framework, the proposed IDI advances compound drought monitoring capabilities and supports more informed climate adaptation and water management strategies. This approach significantly enhances understanding of drought dynamics and provides policymakers and stakeholders with a stronger decision-support tool amid increasing climate variability.