A novel framework for urban flood risk assessment integrating probabilistic hazard occurrence and urban environmental response.

Abstract

Urban flood (UF) risk assessment provides a geographical perspective for addressing threats to urban development and public safety. Although the Hazard-Exposure-Vulnerability (H-E-V) framework is widely used, ambiguous boundaries between Exposure and Vulnerability and unreasonable coupled risk-value distributions may weaken the credibility of spatiotemporal risk representation. To address these issues, this study proposes an optimized urban flood risk (UFR) method oriented toward the hazard-forming process and disaster-bearing environment (DBE) of UF, which clarifies the definitions and quantification of Hazard, Exposure, and Vulnerability, reconstructs the DBE by coupling Exposure and Vulnerability, and applies data transformation to optimize coupled risk-value distributions. Guangzhou, China, was used as the case study. Results showed that (1) compared with the conventional risk formulation based on simple multiplicative coupling, the proposed method achieved a higher density of observed flood locations within high-risk areas under the same area constraint, improving high-risk area identification; (2) the cube-root transformation outperformed other transformations in value range, distribution characteristics, and information entropy, improving distribution rationality by reducing low-value compression, right-skewness, and long-tailed effects while preserving spatial differentiation information; and (3) the identified high-UFR areas were consistent with DBE characteristics revealed by remote-sensing imagery and socioeconomic data, including dense built-up surfaces, flat terrain, proximity to rivers and roads, and concentrated population and economic activity. The proposed H-E-V-based UFR method provides methodological support for high-risk area identification, refined flood management, and sustainable urban development.