Creating a framework for determining the sensitivity of a storm surge model to topographic features and data resolution

Topographic data is required in nearly every hydrological modeling study. While topographic data of various resolutions is available for use, it is not well understood how the resolution affects model results. In this study, it is hypothesized that topographic data resolution can have varying impacts on model results depending on the physical properties of the study area, the model structure, and the processes being modeled. Preliminary work was done for testing this hypothesis with a limited scope in the context of storm surge modeling. A framework was developed that involved creating modifiable synthetic topography files based on real topographic data and modeling the impact of theoretical storms on the synthetic topography. We implemented this framework using 3 m lidar data describing the topography of the Atlantic Coast of the United States. The data was processed to create synthetic topography files of various resolutions and modifiable coastal landforms using GIS and Python. Synthetic storms were created by modifying the trajectory of Hurricane Ike from 2008. Storm surge simulations of the hurricane impacting the synthetic topographies were generated using GeoClaw, a numerical solver for the shallow water equations which utilizes adaptive mesh refinement. Model results were compared to determine the effects of resolution and topographic features. Results show that differences in modeled surge height and timing can both be observed between topographic data resolutions of 3 m and 30 m. This suggests that topographic data resolution has an impact on storm surge model results. This framework could be used in future studies to fully assess the effects of topographic data resolution and features on storm surge model predictions and to determine the resolution requirements to adequately perform storm surge forecasting based on the landform features of specific locations.