Refined impact assessment of rainfall-induced urban flood in a densely populated and built-up watershed

The Liede Watershed, a representative densely populated and built-up watershed in Guangzhou, China. With improved disaster prevention and flood control needs, more refined spatial distribution characteristics of hazard-affected bodies in small-scale urban areas must be analyzed. To address this limitation of current methods for measuring such hazard-affected body distribution, we first employed a coupled model to simulate flooding depths and extents, to derive refined hazard information; two multi-source data types—Points of Interest (POI) and population heatmaps—were adopted, and a convenient method based on combined matrix framework was devised to assess the multifaceted impact of urban flood on assets and population, along with their combined effects. The comprehensive impact results show that high-impact areas primarily consist of commercial zones and residential areas. At 9 pm on weekends—when population vulnerability is the highest in four typical moments—the very high impact zones expand from 0.06 km² (1a) to 0.39 km² (100a) as return periods increase. The two multi-source data elucidate the intricate spatial interior distribution of assets and population, thus mitigating the limitations of fine-grained distribution within small spatial scales. The proposed method facilitates convenient and accurate comprehensive numerical simulations and impact analyses of urban floods, thereby providing a valuable reference for assessments and specific measures of urban flood impacts in urbanized watersheds. • Extended combined matrix framework is adopted to evaluate the comprehensive effect of urban floods. • The multi-source data effectively resolve fine-grained spatial heterogeneity of assets and population at small scales. • The proposed framework provides a convenient and refined assessments of urban flood impacts.