Comparing simplistic versus complex modeling approaches for simulating localized urban flooding

Abstract:

Despite an immense potential, the application of integrated hydrologic and hydraulic models towards flood modeling and forecasting remains challenging without compromising on the spatiotemporal scale and resolution, computational efficiency, and local-scale hydrodynamics. Due to these issues, it is often difficult to simulate hyper-resolution real-time street level flooding in highly urbanized catchments. Alternatively, localized flooding potential can be simulated efficiently by coupling the overflows from a detailed stormwater model into a simplistic flood wave propagation model. This study aims to delineate the best practices in simulating large scale but locally relevant urban flooding for better risk management, green infrastructure planning, and flood forecasting. The hyper resolution integrated model structure (built using ICPR) used in this study can simulate the flow through the surface water, groundwater, and stormwater systems simultaneously. The simplified model (or the rapid inundation model) combines a 1D hydraulic drainage network model (SWMM) and theory of “Storage Cell Method” to rapidly estimate the maximum flood depths and extents by spreading flood volumes to topographic depressions. The effectiveness of these approaches in simulating localized urban flooding is analyzed using the City of Minneapolis, MN as a test bed. Findings from application of the two approaches, including their validation using locally collected water level data will be presented.