Comprehensive assessment of climate model performance in representing drought dynamics at catchment scale

Drought assessment is crucial for climate change adaptation. This study evaluated 10 CMIP6 climate models (MRI-ESM2, EC-EARTH3, CMCC-ESM2, INM-CM4-8, NorESM2-MM, MPI-ESM1.2-HR, INM-CM5, ACCESS-ESM1-5, IPSL-CM6A, and MIROC6) in representing droughts in the Capibaribe River Basin, Brazil. Drought characterization was assessed using the Standardized Precipitation Index (SPI) across multiple time scales (6–48 months) and statistical metrics (R, RB, MSE, RMSE, Kd, Kp). Results indicate that accuracy improves at longer time scales (SPI-24 and SPI-48). MIROC6 performed best overall, while INM-CM5, CMCC-ESM2, and ACCESS-ESM1-5 showed limited precision. Future projections under the Shared Socioeconomic Pathway 2–4.5 (SSP2–4.5) and Shared Socioeconomic Pathway 5–8.5 (SSP5–8.5) scenarios indicate a spatial divergence: coastal areas may face more frequent short-term droughts, while inland regions are likely to experience significantly more prolonged and intense multi-year events. These findings provide an evidence-based foundation for targeted water management to enhance water security against intensifying climate risks.