A Global Probabilistic Framework for Meteorological Drought Risk Assessment Using Self-Calibrating PDSI and Stochastic Simulation

Drought is one of the most consequential natural hazards, with wide-ranging impacts on ecosystems, economies, and societies. Assessing drought risk under current and future climate conditions is essential for adaptation planning and financial risk management, but remains difficult due to the complexity of climate variability and uncertainty in projections. This study develops a global probabilistic framework for estimating meteorological drought return periods that integrates climate data with statistical modeling.The framework combines the self-calibrating Palmer Drought Severity Index, a stochastic weather generator, and generalized extreme value analysis to evaluate drought duration extremes at a 2.5° × 2.5° global resolution. The weather generator reproduces observed variability, persistence, and long-term trends, creating 1,000 synthetic time series per grid cell to enable robust probabilistic assessment. Future projections are derived by scaling the synthetic series across three socioeconomic pathways representing low, medium, and high greenhouse gas emissions (SSP1-2.6, SSP2-4.5, and SSP5-8.5), using temperature adjustment factors from a multi-model climate ensemble. Post-processing corrections are applied to address statistical limitations such as persistence effects and nonlinear behavior in extreme droughts.The results demonstrate broad agreement with climate model ensembles while identifying regions where nonlinear dynamics drive divergences. This framework provides spatially explicit estimates of drought risk, offering a practical tool for decision-makers in climate adaptation, insurance, and financial sectors. More broadly, the study highlights the value of probabilistic approaches that integrate observational statistics with climate projections to strengthen resilience in a warming world.