Flood risk mapping for direct damage to residential buildings in Quebec, Canada

Abstract

This paper presents a probabilistic approach for flood risk assessment in a reach of the Richelieu River, south of Saint-Jean-Sur-Richelieu, Quebec, Canada. The approach is based on a combination of three simple modules: 1) flood frequency analysis (frequency and peak discharge), 2) estimation of inundation depth, and 3) damage and loss estimation. To assess the flood negative impacts, a simple hydraulic model is developed designed to replace and complement the existing 2D model coupled to an existing damage model. By simplifying the spatial coverage of flood calculations, this approach accelerates the computational efficiency enabling a broader set of elements to be combined in a large sample of Monte Carlo simulations applied in the flood risk assessment. The final result is a local scale risk map indicating the expected annual damage to each individual building in the study area useful for flood risk decision making. The analyses show that the approach is particularly powerful for flood risk assessments in areas adjacent to the river for which sufficient data is available.