A Subgrid Correction Scheme for Accurate km‐Scale Coastal Inundation in the Energy Exascale Earth System Model

Abstract

Global Earth system models describe many large‐scale processes, such as sea level changes, that affect the evolution of risk in local‐scale coastal flooding due to extreme events such as hurricane storm surges. However, these models operate at resolutions that are too coarse to resolve coastal ocean tidal and storm surge exchanges across the terrestrial‐ocean interface. This lack of resolution makes it difficult to assess changes in coastal flood risk on decadal timescales. The inclusion of subgrid scale corrections, based on high resolution digital elevation maps, allow coarse grid‐scale variables to be informed by unresolved flow connectivity and frictional dissipation. Here, subgrid scale corrections are introduced into a 2‐D, barotropic configuration of the ocean model component of the Energy Exascale Earth System Model, MPAS‐Ocean, and evaluated in three test cases. The numerical accuracy and convergence of the subgrid model are verified with an analytical test case. A regional model of Buttermilk Bay, MA demonstrates the subgrid model can result in up to a 23× speedup for a given level of accuracy compared to a standard grid‐scale model. Finally, a hindcast of Hurricane Sandy (2012) is used to validate the effectiveness of the subgrid methodology in practical applications within a global model. Together, these results suggest that subgrid corrections can enable coastal exchanges and flooding to be accurately and efficiently represented in global barotropic ocean models.