Sustainable Urban Water Management via AI Surrogate Modeling: A Multi-Dimensional Framework for Drainage Resilience and Microclimatic Assessment

Extreme climate events and rapid urbanization pose severe threats to urban water sustainability and environmental resilience. This study presents a surrogate-assisted hydrological simulation framework designed to support sustainable urban drainage planning and multi-objective scenario exploration. The proposed approach integrates a physics-based hydrodynamic model with a machine learning surrogate (XGBoost) to emulate system responses across diverse design configurations. Under the assumed setup, the framework rapidly evaluates hydrological indicators (peak runoff, flood depth, duration, and extent) alongside microclimatic co-benefits (urban cooling and ventilation) driven by blue-green infrastructure. The surrogate model demonstrates high fidelity R2≥0.93, achieving a speedup factor of 105 to enable sustainable design screening and trade-off analysis between flood mitigation and urban liveability. Furthermore, post-construction water quality baselines demonstrate the framework’s capacity to incorporate holistic environmental metrics. Overall, this research provides a computationally efficient decision-support tool to advance the Sustainable Development Goals (SDGs)—particularly SDG 11 (Sustainable Cities and Communities) and SDG 13 (Climate Action)—by offering an actionable methodology for climate-resilient grey-green infrastructure planning under data-constrained conditions.