Assessing the impact of flood inundation dynamics on an urban environment

This study attempted to examine the complex impact of dynamic inundation process of extreme events on flood hazard assessment (FHA) for the affected urban settings around a local river in New York, USA. Using HEC-RAS 2D, LIDAR DEM, distributed values of surface roughness, and hourly discharges at both ends of the selected reach, we simulated the full inundation process of a 500-year storm event, constructed in terms of the existing largest storm event. We presented flooding status at three flooding moments and quantitatively described the temporal changes of inundation area, depth, and the associated stream power over the entire flood period. Then, we analyzed differences of inundated areas in four classes defined using traditional classification (TC) and process-based classification (PBC). The (static) former was based on the maximum inundation map, while the (dynamic) latter accounted for both inundation depth and duration. We showed that inundated areas in higher classes based on TC were much greater than those in similar classes based on PBC, indicating the significant impact of inundation duration on classification of flood hazard. Next, we investigated the impact of different land use/cover on the difference of inundated areas between the two types of classifications and found that it was complex and displayed no consistent trend from areas surrounding individual buildings (local scale) to large inundated areas (global scale). We emphasize the importance of considering the overall impact of the entire flood processes of an event on future FHA.