Impacts of Elevation Data Spatial Resolution on Two-Dimensional Dam Break Flood Simulation and Consequence Assessment

Abstract

In the United States, there are approximately 84,000 dams including approximately 14,000 dams that are classified as high hazard. Approximately 50% of high hazard dams do not have an emergency action plan (EAP), a document describing potential emergency conditions and potential areas at risk of flooding. A critical data set required for identifying flood risk regions through modeling and simulation is digital elevation models (DEM). These data have become increasingly available at high resolution. The difficulty in utilizing the higher resolution data is that model computation time is increased drastically and becomes, in the case of wide-area (regional) analyses, infeasible to use. The tendency for modelers, therefore, is to use lower resolution data for these model applications. It is clear that when using the lower resolution data that topographic features are not represented as well, but it is not as clear what impact this has on two-dimensional modeling and flood risk estimation. Additionally, there is no rule of thumb as to which resolution should be used. This paper evaluates the impact grid resolution has on estimating the flood risk area resulting from dam failures using two-dimensional models. Results indicate that while flood extent, depths, and flood wave timing are sensitive to grid resolution, socioeconomic metrics such as population at risk and economic loss are less sensitive to simulation grid resolution. This observed socioeconomic insensitivity validates the potential of using coarse resolution simulation as a flood screening tool or in emergency response situations.