Assessment of flood loss in administrative units based on improved vulnerability curves

The vulnerability curve is a key method for assessing flood loss risk, which contributes to the improvement of flood prevention and relief systems. However, existing vulnerability curves are typically developed on a large spatial scale, and intraregional variations in vulnerability are overlooked. This study aims to quantify the spatial heterogeneity of vulnerability within a region and elucidate the impact of vulnerability on flood losses. With Hubei and Hunan provinces in China as case study areas, vulnerability curves are constructed using a mixed-effects model. These curves are then combined with flood intensities across various return periods to estimate flood losses. In addition, the influence of vulnerability and flood intensity on flood losses is analyzed. The results indicate that the mixed-effects model is effective in constructing distinct vulnerability curves for smaller-scale administrative units (e.g. cities) while also assessing the overall vulnerability of the study area, and it achieves high accuracy (R² > 0.75). The disparities in loss rates between cities increase under longer return periods. Furthermore, the variation in vulnerability between cities is the primary factor influencing flood losses for short return periods, whereas differences in hazard intensity have a greater impact on city flood losses during longer return periods. This study provides methods and recommendations for systematic flood risk reduction and proposes pathways for enhancing climate resilience.