ABSTRACT
This chapter aims to present comprehensive techniques regarding continuous large-scale simulation for flood risk studies. The combined effects of elevated water level and wave run-up are the major causes of coastal flooding in the mid-Atlantic states during extreme storm events, i.e., hurricanes. In this study, a state-of-the-art unstructured grid model known as Advanced Circulation (ADCIRC) was used to study the hydrodynamic response in the Western North Atlantic Domain (WNAT), USA, during the superstorm Sandy of 2012, a notable example of hurricanes in this area. The model predictions were validated with the observed tide surges and waves during this storm event. The performance of the ADCIRC model was evaluated by different statistical parameters to assess the model’s ability to reproduce the storm-tide patterns. The overall root mean square error and Nash-Sutcliffe efficiency were in the order 0.2–0.5 m and 0.37–0.8, respectively. Next, a fine-scale 2D hydrodynamic model was implemented by utilizing the boundary forcing from storm surge modeling to produce high-resolution inland flood modeling. Computational efficiency was achieved by implementing the latest high-performance computing and GPU processing. The results of flood inundation inland were validated through the different sources of observed inundation maps and high-water marks. Model performance of inland flood modeling was assessed through several skill metrics. Finally, the high-resolution inundation maps were published through a GIS framework for public dissemination.
