Subgrid modeling of urban flooding incorporating buildings' effects

Abstract

Coastal cities at low-lying elevations face extensive inundation risks when attacked by extreme storm surges or tsunamis. Efficient and accurate modeling of flooding is critical for reducing losses in economic and lives. Therefore, the paradox arises that the computational cost is usually too high for rapid forecasts when considering the complex topography and buildings. The present contribution presents a subgrid-based approach, which combines the subgrid model and a drag force module, to realize efficient modeling of urban flooding at a high resolution. The volume-occupation effects of buildings are incorporated naturally in the subgrid treatment, and the blockage, sheltering, and frictional effects of buildings acting on the flows are considered through a simple, bulk parameterization in the form of drag forces. The model is validated with a laboratory experiment and then applied to an ideal numerical model at a real city scale. The proposed model is compared to three other models using coarse grids, with high-resolution simulation results serving as the benchmark for evaluation. The results demonstrate that the presence of buildings induces noticeable alterations in the distribution of water level and velocity fields within the urban area. The proposed model yields comparable predictions to the benchmark in terms of inundation extent, elevation, and velocity. Furthermore, the model exhibits minor sensitivity to grid size, allowing for a significant reduction in computational time compared to high-resolution simulations. The model achieves a well-balanced combination of high resolution, accuracy, and efficiency for large-scale simulations of coastal urban flooding.