A hindcast model simulated the hydrodynamic conditions from the landfall of Hurricane Isaias in August 2020 along the North Carolina, USA coast (Manchia and Mulligan, 2022). The hydrodynamic model Delft3D was coupled with the spectral wave model SWAN (Simulating WAves Nearshore) to span the coastal region of Onslow Bay and its surrounding continental shelf offshore of Wrightsville Beach. The SWAN model was forced with water level boundary conditions defined by tide gages located at Duck, NC (NOAA station 8651370,
https://tidesandcurrents.noaa.gov/stationhome.html?id=8651370) and at Charleston, South Carolina (SC) (NOAA station 8665530,
https://tidesandcurrents.noaa.gov/stationhome.html?id=8665530). The Simulating WAves till SHore (SWASH) model (Zijlema and others, 2011) was used to simulate wave and water level interactions with the shore. Hourly directional energy spectra produced by Manchia and Mulligan (2022) were output at the offshore boundary of the SWASH domain (water depth = 13.5 m). Ten hours around the peak of the storm were simulated to capture the evolution and maximum water levels of Hurricane Isaias, 23:00 GMT on 3 August 2020 to 09:00 GMT on 4 August 2020. Wave energy was propagated over a 1-Dimensional bathymetric transect located 20 m north of a local pier at Wrightsville Beach, NC. The SWASH computational grid extends 2292.5 m and has a cross-shore horizontal resolution of 2.5 m. The transect was repeated in the alongshore to have 5 grid points, spanning 20 m (delta-y = 4 m) to accommodate directional energy inputs. Input spectral files had 42 frequency bins between f = 0.04:0.4 hertz (Hz) and had a directional resolution of 5 degrees. The wave field was initialized with all current and velocity components set to zero and a mean water level equal to the corresponding verified water level at the tide gauge at Wrightsville Beach, NC (NOAA station 8658163,
https://tidesandcurrents.noaa.gov/stationhome.html?id=8658163). Each simulation was run for a total of 60 minutes with the first 15 minutes allocated for spinup time as the domain reached steady-state. Data were output at a time step of 0.5 seconds (s), but the model internally computed variables at a dynamic time step ranging between 0.04 and 0.08 s, governed by the Courant number. The horizontal viscosity was taken as constant, and a standard k-epsilon model was used in the vertical direction. Default model values were used for the Courant number, the breaking threshold, and in the calculation of bottom friction, which included a constant Manning coefficient (Zijlema and others, 2011). The resulting water level elevations, extracted at the end of each simulation, were output into MATLAB data files. For further information regarding model input generation and visualization of model output topography and bathymetry, refer to Birchler and others (2024).
