The influence of urban morphological changes on pluvial flooding during urban expansion

Urban expansion alters landscape structure and hydrological processes, heightening pluvial flood risks in cities worldwide. This study integrates multi-temporal urban expansion data with the CAFlood cellular automata model to simulate rainfall events with multiple return periods (i.e., 1-in-10, 1-in-100, and 1-in-200 year return periods) across three different urban environments: Shenzhen (China), suburban Houston (United States), and Lausanne (Switzerland). To quantify how evolving urban morphology drives flood volume, we computed key landscape metrics describing impervious extent (percentage of impervious surface), green space shape complexity (Area-weighted Mean Shape Index), green space aggregation (Aggregation Index), and impervious edge complexity (Landscape Shape Index). Their effects on changes in flood volume were analysed with Generalised Additive Models stratified by levels of pre-expansion impervious cover. The results show that decreasing green space aggregation and impervious edge complexity are consistently significant indicators of increased flood volumes, particularly in areas with low pre-expansion impervious cover (<50 %). However, the influence of these two key morphological indicators decline as impervious coverage increases, with total imperviousness gradually becoming the dominant factor. Sensitivity experiments with synthetic urban fabrics confirm that maintaining compact, well-connected green spaces can limit the increase in flood volume even with when impervious surfaces expand equally. By explicitly linking the morphological changes in urban form to pluvial flood response across diverse climates and urbanisation context, this study moves beyond static approaches and provides actionable insights for planning flood-resilient urban growth.