To address the challenges posed by extreme rainfall, this study proposes a spatial optimization model for deploying low-impact development (LID) facilities, with a university campus in Lianyungang, China, as the case study. The model minimizes the construction cost under the rigid constraint that the runoff control rate must be ≥70%. The Storm Water Management Model (SWMM) is used as the simulation engine, and the active covariance matrix adaptation evolution strategy (active CMA-ES) algorithm enables intelligent spatial optimization across 2-, 5-, 10-, and 20-year return periods. The cost-effective configurations are further simulated using the coupled SWMM–synergizing high-performance hazard simulation with data flow (SynxFlow) hydrodynamic model. Results show that, compared to the scenario without LID measures, the optimized solutions can reduce the peak runoff by 58.08%–63.83%, reduce the maximum inundation depth from 0.84 to 0.41 m, and lower the node ponding volume by over 63%. More importantly, the framework reveals a dynamic adaptation pattern in the composition of LID facilities: Under low-intensity rainfall, the solution is dominated by cost-effective facilities (PP), accounting for 73% of the total area; as the rainfall intensity increases to the 20-year return period, the solution automatically adjusts to a balanced structure with the collaborative deployment of multiple facility types, where the proportion of PP decreases to 38%. This “spatial data-coupled modeling-intelligent optimization” framework demonstrates strong applicability for flood risk mitigation and offers a practical and cost-efficient strategy for LID deployment in sponge city development.
Xu et al. (Fri,) studied this question.