Socioeconomic Risk Assessment of Rainstorm Disasters in Shenzhen Based on Vulnerability Curves

Rapid expansion of megacities and the concentration of high-density assets have amplified the economic impacts of urban rainstorms. To quantify these impacts, a socioeconomic risk assessment model for urban rainstorm disasters was developed using precipitation observations from national and regional meteorological stations in Shenzhen, records of rainstorm-related disasters, and socioeconomic data such as fixed-asset investment for the period 2007–2021. Hazard was characterized by the integrated intensity of individual rainstorm events, economic exposure by fixed capital stock, and the degree of loss by the direct economic loss rate. Economic vulnerability curves were fitted at both the city and district levels. The model was subsequently applied to the rainstorm events of 6–8 June 2022 and 7–8 September 2023 for risk classification and loss validation. The results show that both the city- and district-level vulnerability curves, which describe changes in the direct economic loss rate with increasing hazard intensity, exhibit a typical S-shaped pattern with good goodness of fit. Simulated direct economic losses were generally of the same order of magnitude as the reported losses, indicating that the model can reasonably reproduce observed disaster losses. District-level simulations outperformed city-level simulations, although an overall tendency toward overestimation remained. Case studies of the two rainstorm events further showed that high-risk areas encompassed the vast majority of reported disaster locations, demonstrating the model’s capability to anticipate the spatial distribution of economic risk. For the 2023 event, the district-level estimate of direct economic losses was 1.901 billion yuan, closer to the reported value of 1.262 billion yuan than the city-level estimate of 2.220 billion yuan. The proposed model provides a quantitative basis for urban rainstorm socioeconomic risk zoning and the determination of risk-warning thresholds.