Urban Pluvial Flood Simulation with Virtual Drainage Networks

Urban pluvial flooding is escalating as a critical hazard, yet high-fidelity modeling remains hindered by the widespread lack of detailed drainage network data. To address this issue we propose a virtual drainage generation framework that explicitly incorporates off-road secondary pipes , often omitted — in conventional road-aligned approaches—to improve flood simulation accuracy in data-scarce areas. The framework is evaluated on the Tongji University campus using SERGHEI-SWMM, a parallel 2D-1D urban flood model coupling a GPU-portable 2D shallow water equations (SWE) solver with a 1D drainage module. Simulations compare the real drainage system against virtual networks with and without off-road pipes. The results highlight the importance of off-road secondary pipes in providing additional spatial coverage of the drainage system, which facilitates drainage in off-road depressions. Implementing the off-road virtual drainage system decreases mean water depth errors by 28.1% when actual pipe data is unavailable. These findings underscore the reliability and practical utility of off-road pipes in virtual drainage networks for urban pluvial flood modeling particularly in small-scale or data-scarce urban environments , .