Improving Coastal Flood Risk Assessment for Large Inland Lakes

Abstract:

The flood hazard due to coastal storm surge has been widely studied using a variety of modeling approaches. However, most of these studies focus on ocean coastal areas, where very few studies have been performed for large inland lakes. Nevertheless, lake coastal flooding is becoming more frequent in recent years, causing catastrophic property damage and loss of life. To address this knowledge gap, this study takes the Great Lakes coastal region as an example, to analyze the impacts of different modeling configurations on coastal flood risk assessment. The 3D unstructured-grid Finite Volume Community Ocean Model (FVCOM) is applied in this study. Four different modeling setups are evaluated: (i) grids of 50m – 500m resolution are generated for the lake area for storm surge modeling, nearshore water level is compared to the land topographic data for defining flooding extension; (ii) in addition to previous configuration, coupled wave-current processes are simulated; (iii) grids are extended to coastal zones with 10m – 50m resolution, only storm surge is modeled, flooding extensions are defined by tracing the wet-dry boundaries; (iv) coupled wave-current processes are included for simulations using the extended grids. Simulated water level is compared with measurements at shorelines to assess the accuracy of different model settings. Moreover, simulated flooding extent is compared between approaches. The results of study provide direct evidence on the importance of including different processes in storm surge simulation for coastal zones of large inland lakes. In addition, comparing with the Digital Flood Insurance Rate Maps (DFIRMs) of the Federal Emergency Management Agency (FEMA), we could make suggestions for improving coastal flooding analysis on Great Lakes.