Hydrodynamic Simulation of the Columbia River, Hanford Reach, 1940--2004 [electronic resource].
- Washington, D.C. : United States. Dept. of Energy, 2005.
Oak Ridge, Tenn. : Distributed by the Office of Scientific and Technical Information, U.S. Dept. of Energy.
- Additional Creators:
- Pacific Northwest National Laboratory (U.S.), United States. Department of Energy, and United States. Department of Energy. Office of Scientific and Technical Information
- Restrictions on Access:
- Free-to-read Unrestricted online access
- Many hydrological and biological problems in the Columbia River corridor through the Hanford Site require estimates of river stage (water surface elevation) or river flow and velocity. Systematic collection of river stage data at locations in the Hanford Reach began in 1991, but many environmental projects need river stage information at unmeasured locations or over longer time periods. The Modular Aquatic Simulation System 1D (MASS1), a one-dimensional, unsteady hydrodynamic and water quality model, was used to simulate the Columbia River from Priest Rapids Dam to McNary Dam from 1940 to 2004, providing estimates of water surface elevation, volumetric flow rate, and flow velocity at 161 locations on the Hanford Reach. The primary input data were bathymetric/topographic cross sections of the Columbia River channel, flow rates at Priest Rapids Dam, and stage at McNary Dam. Other inputs included Yakima River and Snake River inflows. Available flow data at a gaging station just below Priest Rapids Dam was mean daily flow from 1940 to 1986 and hourly thereafter. McNary dam was completed in 1957, and hourly stage data are available beginning in 1975. MASS1 was run at an hourly timestep and calibrated and tested using 1991--2004 river stage data from six Hanford Reach locations (areas 100B, 100N, 100D, 100H, 100F, and 300). Manning's roughness coefficient in the Reach above each river recorder location was adjusted using an automated genetic algorithm and gradient search technique in three separate calibrations, corresponding to different data subsets, with minimization of mean absolute error as the objective. The primary calibration was based on 1999, a representative year, and included all locations. The first alternative calibration also used all locations but was limited in time to a high-flow period during spring and early summer of 1997. The second alternative calibration was based on 1999 and included only 300 Area stage data. Model goodness-of-fit for all years with data was high in the primary calibration and indicated little bias caused by selecting 1999. The alternative calibrations led to improved goodness-of-fit for their limited time and locations, but degraded goodness-of-fit overall. Overall, the simulations were very accurate and even highlighted some probable data problems, as evidenced by systematic shifts in the data. Further improvements in simulating the historic period would depend on correcting these inferred data problems. For all years and locations, the mean absolute error in the primary calibration was 14.8 cm, the mean error was 1 mm, and model efficiency was 0.988. The MASS1 output for 1940--2004 can be used to reconstruct historical river elevations at Hanford or to build scenarios of future river elevations for solving environmental problems such as groundwater-river interaction or fish habitat inventories. Model output and additional processing services are available from the authors. Longer-term scenarios extending more than a few decades from now should also consider the impacts of climate change and reservoir operation change. Once defined, these impacts could be used to drive new simulations with MASS1.
- Report Numbers:
- E 1.99:pnnl-15226
- Other Subject(s):
- Published through SciTech Connect.
Richmond, Marshall C.; Waichler, Scott R.; Perkins, William A.
- Funding Information:
View MARC record | catkey: 14410127