Comparative Evaluation of two Modeling Approaches of Urban Flooding in Ottawa, Canada

Abstract

Urbanization and climate change have heightened urban flooding, necessitating stormwater modeling approaches tailored to urban contexts. This study compares two modeling approaches using PCSWMM for an urban catchment in Ottawa, Canada: (1) a topographic model, which simulates drainage patterns based on a 0.25-meter DEM, and (2) a lot-level model, which represents idealized drainage by routing runoff directly to roads. While the topographic approach reflects conventional subcatchment delineation, the lot-level approach is designed to evaluate impacts of property-level actions on flooding. Both models, provided by the City of Ottawa, were calibrated using flow and precipitation data for 11 storm events and validated for 6 events. Calibration showed correlations ranging from 0.61 to 0.97 (average 0.84) and KGEs averaging 0.49, while validation achieved correlations of 0.84–0.86 and KGEs between 0.42 and 0.34. A 69-mm storm event was simulated to assess flood distribution differences. The topographic model identified greater backyard flooding due to terrain-driven routing, whereas the lot-level model predicted higher street inundation caused by direct routing to roads. While no surface water level data were available for model calibration, results reflect relative differences due to routing assumptions. This analysis informs targeted flood management strategies by aligning modeling methodologies with specific mitigation objectives.