Simulating Compound Flooding with an Efficient High-Resolution Sub-Grid Model

Abstract:

We present a high-resolution sub-grid hydrodynamic inundation model MOI (also known as ELCIRC-sub) that aims to efficiently simulate compound flooding induced by coastal storm surge, extreme riverine flow, and heavy urban rainfall. The model is coupled with a coastal storm surge model, a riverine flow model, and an urban runoff model. MOI uses a high-resolution sub-grid (e.g. 1 m) and a coarse resolution finite volume computational base grid (e.g. 200 m). The momentum and mass fluxes calculated by the coarser base grid model are coupled with the sub-grid so that running a full-blown high-resolution base model is not required. Also, MOI uses MPI (Message Passing Interface) parallel computing to support regional scale simulations and an efficient non-linear solver to improve the accuracy of the wetting-and-drying processes. The model has been validated in the following three hindcast studies: (1) 2012 Hurricane Sandy in New York City: hit rate > 85% compared with USGS observed flood extent and averaged error < 0.2 m compared with USGS observed water level; (2) 2017 Hurricane Harvey in Texas San Jacinto river basin: hit rate > 95% and threat score > 0.85 compared with USGS observed flood extent. (3) 2016 flooding in Kumamoto, Japan: averaged error < 0.3 m compared to observed high water marks. In addition to its high accuracy, MOI runs efficiently with 100x real-time speedup on multi-core HPC clusters. We demonstrate that MOI would be suitable for both real-time live flood prediction and flood risk assessment.