A dynamic framework for assessing water quality resilience in urban rivers and its application to low-impact development.

Urban rivers are facing severe water quality fluctuations and prolonged recovery challenges due to stormwater runoff pollution amid rapid urbanization and climate change. This study proposes a dynamic water quality resilience assessment framework by coupling the Storm Water Management Model (SWMM) with a completely mixed river model. Grounded in system performance curve theory, the framework integrates a failure index with nonlinear penalty coefficients to quantify exceedance magnitude, duration, and the full response-recovery trajectory, from resistance and absorption to restoration, thereby overcoming the limitations of traditional static indicators. Applied to the Xixiang River, a typical rain-fed urban river in Shenzhen, the framework revealed pronounced seasonal and pollutant-specific vulnerabilities under baseline conditions. Multi-scenario simulations demonstrated that Low-impact Development (LID) significantly enhances water quality resilience through source-load reduction, peak-flow attenuation, and hydrograph smoothing. At 80% LID coverage under a 1-year return period storm, loads of Chemical Oxygen Demand (CODcr), Total Phosphorus (TP), Total Nitrogen (TN), and Ammonia Nitrogen (NH3-N) decreased by 60.33-78.11%, with markedly improved resilience and faster recovery times across wet and dry seasons. The results highlight the critical role of source control in bolstering urban river resilience under climate change, while revealing LID's limitations for extreme events and dissolved pollutants. The proposed framework provides a robust and transferable tool for resilience-oriented urban water management and supports integrated "gray-green-blue" strategies for sustainable and climate-resilient urban aquatic ecosystems.