Developing Practical Green-Grey Solutions Based on Critical Node Identification for Stormwater Management

Urbanization has exacerbated challenges faced by urban watersheds, including increased impervious surfaces, deteriorating water quality, and heightened flood risks. Previous research has extensively employed the Genetic Algorithm (GA)  to optimize urban grey-green infrastructure (GGI), primarily focusing on preventing system-wide overflow during design storm events. However, the high costs associated with these solutions have often hindered their implementation. This study proposes a practical approach to enhance urban stormwater management by prioritizing interventions at critical locations within watersheds. A multi-index fuzzy comprehensive evaluation (MFCE) model was developed to identify critical nodes in the drainage network based on hazard (overflow volume and duration), topological characteristics (degree and Katz centrality), and vulnerability (peak hour traffic flow). Problematic segments within the drainage network, including those with adverse slopes, mismatched pipe diameters, and ground depressions, were identified using a combination of SWMM simulations and graph-based analyses. Subsequently, the Genetic Algorithm (GA) was employed to optimize the design and placement of grey-green infrastructure solutions, subject to the constraint of preventing overflow at these critical nodes during design storm events. A case study in Guangzhou, China, demonstrated the efficacy of this approach. The optimized grey-green infrastructure system significantly reduced budgetary costs and peak flow compared to traditional grey infrastructure systems, while enhancing flood control and improving the overall resilience of the urban watershed.