Enhancing drought monitoring through regional adaptation: Performance and calibration of drought indices across varied climatic zones of Iran

Iran. This study evaluates the performance of various drought indices, including SPEI (Standardized Precipitation Evapotranspiration Index), Standardized Soil Moisture Index of the top two layers (SSI 1 and SSI 2 ), and the Multivariate Standardized Drought Indices (MSDI 1 (P&ET ref ), MSDI 2 (P&SM 1 ), and MSDI 3 (P&SM 2 )) models, across six distinct climatic zones using data from 30 basins with 621 gridded points (1979–2022). The analysis covers three time scales—1, 3, and 12 ∼ months—and assesses the drought characteristics and criteria in diverse climate regions. The MSDI models exhibited superior performance across all climatic zones, achieving an overall precision rate of 85 % and consistently outperforming the SPEI and SSI models in both short-term (1- and 3-month) and long-term (12-month) drought predictions. In coastal wet and mountain regions, the MSDI models demonstrated exceptional precision rates of 90 % and 85 %, respectively, with robust Taylor skill scores of 0.92 and 0.89, significantly surpassing the accuracy of the SPEI and SSI models. In semi desert and desert regions, the MSDI models maintained a precision rate of 77 %, with a slight decline at the 12-month scale. Despite this decrease, they continued to outperform the SPEI and SSI models, particularly in short-term (3-month) drought assessments. These findings underscore the necessity of selecting and calibrating drought indices to enhance monitoring accuracy, with the MSDI models proving particularly reliable in semi-desert and mountainous regions. The study advocates for region-specific drought indices to better capture local climatic variations and emphasizes the importance of improved model calibration in regions exhibiting lower performance. Policymakers are urged to implement tailored drought management strategies to enhance water resource sustainability, strengthen agricultural resilience, and mitigate the adverse impacts of drought. Further research is essential to refine these models and integrate advanced methodologies, such as machine learning (ML), to enhance drought prediction accuracy and support climate adaptation efforts. • Drought trends and uncertainties were assessed across Iran’s climatic regions using SPEI, SSI, and MSDI indices. • MSDI models showed 85% precision, excelling in coastal wet (90%) and mountain (85%) regions. • SPEI and SSI showed high variability in coastal desert areas, with uncertainty (U 95 %) of 3.26, requiring better calibration. • Short-term (3-month) drought predictions were most reliable, while long-term (12-month) ones declined, especially in deserts. • The study highlights the need for region-specific models and ML-based calibration to improve drought monitoring and resilience.