Hydrodynamic modeling of the mouth of the Columbia River, Oregon and Washington, 2013

A hydrodynamic model of the mouth of the Columbia River (MCR) was constructed using Delft3D, an open-source software package used to solve the unsteady shallow water equations to simulate water motion due to tides, waves, wind, and buoyancy effects (Lesser and others, 2004). The model application for this study was adapted from Elias and others (2012) who provide a detailed description of the model formulations and setup. Changes were applied to the existing model grid and boundary conditions, as described below, to simulate hydrodynamics in quasi-real time for the time period of the RIVET II field experiment between May 9 and June 15, 2013. A curvilinear grid consisting of approximately 68,500 grid cells was used to simulate hydrodynamics throughout the MCR and adjacent coast and estuary. Ten equally spaced vertical sigma layers were used to simulate 3D effects within the model domain. The grid was aligned with coastal engineering structures including the 3 primary jetties as well as several training dikes in the vicinity of the MCR. Flow through the structures was limited in the model using thin dams (no transmission). The spatial extent of the grid was expanded in the ocean, extending roughly 150 and 100 km north and south of the MCR, respectively.

The model bathymetry was derived from recent datasets including the swath bathymetry collected as part of this study (Gelfenbaum and others, 2015). Additional bathymetric data used in the model setup included swath bathymetry collected by NOAA between 2007 and 2009 (online data available at https://maps.ngdc.noaa.gov/viewers/bathymetry/) and unpublished single-beam bathymetry collected in 2004 and 2012 by the U.S. Army Corps of Engineers. A previously published DEM of the lower river (http://www.estuarypartnership.org/lower-columbia-digital-terrain-model) was used for the tidal river between the Astoria Bridge and upstream boundary. A regional digital terrain model (Love and others, 2012) was used in areas where more-recent datasets were not available. The source bathymetry data were converted to a common horizontal datum (NAD83, CORS96) and the land-based North American Vertical Datum of 1988 (NAVD88), projected in the Cartesian UTM Zone, 10, meters coordinate system, and interpolated onto the computational domain of the model. Deep areas associated with the Astoria submarine canyon were removed from the model bathymetry to improve stability along the oceanic boundaries.

Oceanic boundaries of the MCR model were forced using astronomic tidal constituents initially derived from the TPXO 7.2 global tide model (Egbert and Erofeeva, 2002) and adjusted during calibration. An initial vertical offset of 1.15 m (positive up) derived from NOAA VDatum (version 3.2; http://vdatum.noaa.gov/) was applied at the oceanic boundary to account for the difference between local mean sea level and NAVD88. Water levels in the MCR and estuary are influenced by coastal processes such as regional upwelling and downwelling events that induce variations at subtidal frequencies (MacMahan, 2016). Oceanic subtidal variations were imposed as at the oceanic open boundary as a time-varying correction to the astronomic tides. The subtidal time-series was derived from observations of water levels at NOAA stations 9440910 (Toke Point, WA) and 9437540 (Garibaldi, OR). Water level time-series from the two stations were obtained for 2013 and a low-pass filter was applied to remove fluctuations at tidal frequencies. The low pass filtered values from both stations were highly correlated and an average for the two stations was applied to the oceanic model boundaries. The landward boundary was forced with a time-series of river discharge measured at 30 min intervals at the USGS river gauge 14246900 (http://waterdata.usgs.gov/usa/nwis/uv?site_no=14246900) located at the Beaver Army Terminal near Quincy, Oregon.

Water level dynamics in the MCR and estuary are controlled by various processes including tides, river discharge, atmospheric forcing, and to a lesser degree hydropower operations (Jay and others, 2015). Analysis of long-term records collected in the study area (Jay and others, 2015) suggest that tides dominate lower estuary, accounting for 60-70 percent of the total water level variance. Therefore, accurate modeling of tidal propagation is essential to simulate hydrodynamics in the MCR. Tidal propagation into inlets like the Columbia River is modified by bed friction, topographic funneling, and opposing river flow. While river flow and topographic variability are well represented in the model by recent high-resolution inputs, bottom roughness is poorly constrained. Previous modeling studies of the MCR and estuary found that tidal propagation is sensitive to bed roughness (Elias and others, 2012). Tidal propagation in the estuary was calibrated using the Manning roughness coefficient as the primary calibration parameter. Simulated water levels were compared against time-series measurements of observed water levels from 4 locations throughout the study area for the time period spanning the RIVET II experiment simulations (May 9, 2013 to June 15, 2013). The sensitivity of tidal propagation to bottom roughness was examined for a range of Manning roughness coefficients between 0.0202 and 0.0260 in a series of 6 otherwise identical simulations. A Manning roughness coefficient of 0.0218 best-represented mean water levels and variance throughout the domain. Additional adjustments to the oceanic boundary conditions were applied to the model using this roughness coefficient. First, the mean water surface elevation applied at the oceanic boundary, or so-called A0 astronomic constituent, was adjusted to account for the offset between mean sea level and the model bathymetry datum, NAVD88. The original A0 estimated from VDatum was adjusted to match water level observations at Astoria. Next, harmonic tidal analysis was performed (Pawlowicz and others, 2002) on the simulated and observed water levels over the analysis time frame. Tidal constituents applied at the boundary were corrected using measurements obtained at Astoria. The final calibrated model accurately simulated water levels at each of the 4 observation stations with total RMSE less than 10 cm.

Edited metadata to add keywords section with USGS persistent identifier as theme keyword. No data were changed.