Authors
Laura McGee (CIRES,NOAA/PSL), Matthew Newman (NOAA/PSL), Sang-Ik Shin (CIRES,NOAA/PSL), John Albers (CIRES,NOAA/PSL), Paige Hovenga (CIRES,NOAA/PSL), John Callahan (Ocean Associates, Inc.), Matthew Conlin (Ocean Associates, Inc.), Saeed Moghimi (NOAA/CSDL), Felicio Cassalho (NOAA/CSDL), Y. Joseph Zhang (VIMS)
Abstract
High tide flooding affects communities across the United States. The current NOAA High Tide Flooding Outlook uses a statistical model that only predicts flooding at tide gauge locations. A climate model-driven dynamical model could instead predict high tide flooding at much finer resolutions along the coast, but currently no suitable model exists for this purpose. This study tests the ability of a SCHISM 2D model to predict high tide flooding events over the period 1993-2018, given reanalysis data as forcing. Results are compared to an ADCIRC+SWAN model and one year of a 3D SCHISM model. The SCHISM 2D model outperforms both the ADCIRC+SWAN model and 3D SCHISM model for all daily maximums. High tide flooding events are predicted more accurately in the Southeast and Gulf coasts, but with mixed results in the Northeast. An investigation into whether tidal forcing needs to be added into the model boundary conditions or can be added in post-processing shows that while adding the tides in post tends to underpredict daily maximum water level variance by up to 10%, there is overall little difference when predicting the high tide flooding events.