Modeled surface waves from winds in South San Francisco Bay
Dates
Publication Date
2024-03-14
Start Date
2021-05-30
End Date
2022-05-19
Citation
WinklerPrins, L.T., and Stevens, A.W., 2024, Modeled surface waves from winds in South San Francisco Bay: U.S. Geological Survey data release, https://doi.org/10.5066/P9QH0GU5.
Summary
A model application using the phase-averaged wave model SWAN (in Delft3D) was developed to simulate wind waves in South San Francisco Bay, California, between 30 May 2021 and 19 May 2022. This data release describes the development of the model application, provides input files, and includes output from the model simulations in netCDF format. Model Application The model application included two domains (Fig. 1) that were 1-way coupled. The coarse overall model domain (wsfb_g1.grd) included the coastal ocean across the entire San Francisco Sacramento/San Joaquin Bay-Delta region was forced along the oceanic boundaries with measured time-varying, spatially uniform wave parameters derived from the Coastal Data Information Program (CDIP) [...]
Summary
A model application using the phase-averaged wave model SWAN (in Delft3D) was developed to simulate wind waves in South San Francisco Bay, California, between 30 May 2021 and 19 May 2022. This data release describes the development of the model application, provides input files, and includes output from the model simulations in netCDF format.
Model Application
The model application included two domains (Fig. 1) that were 1-way coupled. The coarse overall model domain (wsfb_g1.grd) included the coastal ocean across the entire San Francisco Sacramento/San Joaquin Bay-Delta region was forced along the oceanic boundaries with measured time-varying, spatially uniform wave parameters derived from the Coastal Data Information Program (CDIP) wave buoy station 029 (Point Reyes), incorporated into the wavecon.wsfb files. The nested model (wsfb_g2.grd) covered south San Francisco Bay. The model was forced at the free surface with simulated wind data from the High Resolution Rapid Refresh (HRRR) atmospheric model produced by NOAA. Water level data to simulate water level variations across the model domain were provided from NOAA Station 9414523 (Redwood City, California) and bathymetry data were compiled from Fregoso et al. (2017) and Fregoso et al. (2020). Water density varied between 1021 and 1023 kg/m3 depending on the season to reflect brackish conditions; a Collins bottom-friction formulation was used with coefficient 0.02; Van der Westhuysen whitecapping was used for better performance in wind-dominated systems as South San Francisco Bay is largely protected from ocean-originating waves. The model inputs were split into 4 time intervals (summer, fall, winter, spring) and run sequentially to simulate approximately 1 year (May 30 2021 to May 19 2022).
Figure 1. Map showing the spatial extents of the two model domains (left) and example of simulated significant wave height in South San Francisco Bay (right). BMV is the station name where we compared modeled wave properties to observations from summer 2021 and winter 2021-2022. The NOAA station indicates the location of our water level data.
Model Validation of the Wave Field
Simulated significant wave heights (Hs) were compared against in-situ observations derived from a Nortek Vector acoustic Doppler velocimeter installed at a platform called “BMV” from 07 June to 11 August 2021 (“summer,” with an instrument swap on July 8th) and 23 November 2021 to 31 January 2022 (“winter”), with further details published in Ferreira et al., (2023). This sensor was installed just subtidally in the mudflats about a kilometer offshore of the eastern shore of south San Francisco Bay (see point BMV in Fig. 1). Timeseries comparisons from both seasons are visible in Fig. 2.
Figure 2. Timeseries of observed significant wave heights (blue) versus modeled (orange dotted) in the summer (top) and winter (bottom) seasons, at site BMV.
In the summer, the model recreated the daily sea breeze-driven wave field. For the summer season, RMSE of simulated significant wave heights was 0.08 m with a modeled seasonal-mean significant wave height of 0.22 m and bias of -0.031. In the winter, RMSE of simulated significant wave heights was 0.07 m with a modeled seasonal-mean significant wave height of 0.10 m and a bias of -0.035 m. Despite proportionally high RMSE values to modal wave heights, there was decent agreement in the distribution of wave heights across the seasons (Fig. 3), which are ultimately critical towards average and cumulative wave power metrics relevant for marsh-edge erosion, which this model was developed for.
Figure 3. pdfs showing comparisons of observed and simulated wave height distributions for the summer (top panel) and winter (bottom panel) time periods.
Digital Data
Model input files compatible with windows executable of Delft3D4 version 4.04.01 and wave model SWAN (version 40.72ABCDE) are provided in the zip archives listed below. Model inputs and outputs were split into 4 time periods that are season-specific ("season_1_summer_io.zip","season_2_fall_io.zip","season_3_winter_io.zip", and "season_4_spring_io.zip") and run sequentially to simulate approximately 1 year.
In each, the following files are inputs to the model:
hrrr[year]_[season]_3km.amu and hrrr[year]_[season]_3km.amv (HRRR wind input files)
wave_sensors_[seasons].loc (northing/easting locations of BMV and a second, more nearshore sensor station, used in the journal article)
And the following files are outputs from the model:
wavh-wsfb-wsfb_g1.nc & wavh-wsfb-wsfb_g2.nc are timeseries output at points specified in the .loc file on the fine and coarse grids, respectively.
wavm-wsfb-wsfb_g1.nc & wavm-wsfb-wsfb_g2.nc, and the similarly-named .dat files are map output in across the fine and coarse grids, respectively.
References
Ferreira, J. C. T., Lacy, J. R., Mcgill, S. C., WinklerPrins, L. T., Nowacki, D. J., Stevens, A. W., & Tan, A. C. (2023). Hydrodynamic and sediment transport data from Whale’s Tail marsh and adjacent waters in South San Francisco Bay, California 2021-2022. U.S. Geological Survey. https://doi.org/10.5066/P972R6AW
Fregoso, T. A., Jaffe, B. E., & Foxgrover, A. C. (2020). High-resolution (1 m) digital elevation model (DEM) of San Francisco Bay, California, created using bathymetry data collected between 1999 and 2020 (ver. 2.0, July 2021). https://doi.org/10.5066/P9TJTS8M
Fregoso, T. A., Wang, R.-F., Alteljevich, E., & Jaffe, B. E. (2017). San Francisco Bay Delta Bathymetric/Topographic digital elevation model (DEM)—2016 SF Bay Delta DEM 10-m. https://doi.org/10.5066/F7GH9G27 → A model application using the phase-averaged wave model SWAN was developed to simulate wind waves in South San Francisco Bay, California, between May 30, 2021, and May 19, 2022. This data release describes the development of the model application, provides input files suitable for running the model using Delft3D version 4.04.01, and includes output from the model simulations in netCDF format.