Spatiotemporal disparities in flood exposure during transformations from natural areas to settlements

Accelerated urbanization and climate change are sharply intensifying flood hazards globally. Consequently, dense socioeconomic assets are increasingly expanding into high-risk flood zones. However, the interactive mechanisms and relative contributions of natural versus anthropogenic drivers remain poorly quantified, hindering tailored risk management. Therefore, this study presents a multidimensional attribution analysis of the changing characteristics of flood exposure areas in China under varying hazard levels during 1985–2015 (a timeframe marked by China's most rapid economic development). Through the integration of settlement footprint data, flood hazard maps, precipitation indices, and economic indicators, we aim to elucidate natural and economic drivers contributing to these changes. The results reveal that the settlement areas of China have demonstrated sustained upward growth, increasing from 100,806.57 km 2 in area in 1985 to 270,440.24 km 2 in 2015 (+168%). Notably, the areas classified under high and very high flood hazard levels underwent the most rapid expansion, with their respective areas increasing by 219% and 226% over the study period. Subsequently, using the Shapley Additive exPlanations method, we demonstrate that economic factors (i.e., gross domestic product) dominantly drive changes in flood exposure areas over time, whereas natural factors (i.e., precipitation) emerge as more spatially prominent. Furthermore, cluster analyses reveal distinct regional patterns in flood exposure, precipitation concentration, and economic development. These findings highlight the importance of adopting regionally tailored flood risk management strategies in settlement planning that factor the unique characteristics of each region.