Modeling and prediction of climate change impacts on water resources vulnerability: A multi-model approach

In a rapidly changing world, uncontrolled climate change worsens water scarcity disrupting hydrological cycles and hindering sustainable development. Addressing water resources vulnerability requires holistic approaches to better understand complex systems, mitigate risks from changing weather patterns, and develop adaptive water management strategies. In this study, we modeled climate change impacts on water resource vulnerability using machine learning (ML) and SWAT model based on CMIP6 Global Climate Model (GCMs) under Shared Socioeconomic Pathway (SSP). Six ML models were evaluated to reliably predict hydroclimatic events; Extremely Randomised Trees (ERT) and Categorical Boosting (CatBoost) performed best for simulating ensemble climate interactions. The statistical indicators confirmed model reliability reducing input uncertainties with bias-corrected datasets. The ensemble SWAT model simulation showed a good agreement between simulated and observed values (R 2 = 93 %, NSE = 91 %, and PBIAS = -1.08 %) for calibration and (R 2 = 94 %, NSE = 93 %, and PBIAS = −2.32 %) for validation periods. Furthermore, we developed a novel Hydrologic Vulnerability Index (HVI) framework based on water balance components to quantify watershed vulnerability dynamics across baseline and future scenarios. The HVI ranged from low to extreme, with maximum lower values (54.03 %) observed at baseline, indicating resilience to hydrological stress, and higher values indicating severe vulnerability (43.45 %) at SSP245, indicating extreme drought conditions. The HVI framework integrates climate projections with actionable insights, offering a comprehensive approach to sustainable water management, adaptive infrastructure, and targeted interventions. Hence, innovative policies are critical to address extreme HVIs ensuring resilience against water scarcity and ecosystem degradation. This study underscores the importance of coupling data-driven hydrological analysis with climate responsiveness for effective watershed and environmental sustainability. These results demonstrate the importance of integrating various perspectives and strategies to address both short- and long-term climatic problems, by employing adaptive management practices to ensure sufficient water and ecosystem resilience. • Bias-corrected ensemble model shows simulation improvements in SWAT model. • ERT and CatBoost found most effective machine learning models. • A framework for assessing hydrologic vulnerability was proposed. • Effective climate strategies are needed to tackle watershed vulnerability. • The study highlights the importance of scenario-based water management.