Urban flood intensification by storm clustering: The impact of rainstorm ‘train effect’

Abstract

<p>This study investigates the impact of rainstorm ‘train effect’ on spatiotemporal inundation of the resultant flooding, based on the 2023.9.7 rainstorm event in Hong Kong, characterised by a pronounced ‘train effect’ with several consecutive rainfall peaks. A two-dimensional (2D) flood hydrodynamic model was employed to simulate the rainstorm-driven flooding process, integrating high-resolution terrain, land use, and radar-based rainfall data. Model validation against 80 observed inundation locations achieved an 86 % coverage rate, confirming its reliability in reproducing flood dynamics at key inundation sites. The influence of the ‘train effect’ of rainstorm on flood response characteristics was assessed based on key parameters including catchment slope, percentage of impermeable area, its spatial aggregation index and rainstorm spatial variability index. Results indicate that rainstorm with the ‘train effect’ significantly increase flooding hazards, particularly in densely urbanised regions. Catchments with flatter terrain and clustered impermeable surfaces are especially prone to severe inundation due to prolonged water accumulation during successive rainfall peaks. Furthermore, upstream-concentrated rainfall notably amplifies the inundation hazard associated with the ‘train effect’ compared to downstream-concentrated rainfall. These findings emphasise the critical role of rainfall pattern in flood hazard management and provide a scientific basis for enhancing urban flood mitigation strategies and resilience planning under extreme weather conditions.<br></p>