USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2008–5093

Back to Table of Contents

Previous Investigations

Since 1981, four surface-water models have been developed for the Coeur d’Alene River from Enaville and Pinehurst to the lake near Harrison. Each model was developed for a unique purpose; therefore, each poses its own drawbacks and limitations.

FEMA Model (1981)

In 1981, the Federal Emergency Management Agency (FEMA) completed the first model of the lower Coeur d’Alene River for a flood insurance study (Federal Emergency Management Agency, 1979, and 1982). This model delineated the 100-year floodplain. The 100-year flood has a 1 percent chance of occurring in a given year. Discharge for the 100-year flood is 69,000 ft³/s at the Cataldo gaging station (12413500) (updated from Berenbrock [2002] through 2007 [n = 76]). For the area between Coeur d’Alene Lake and Interstate Highway 90, elevations and delineation for the 100-year flood were based on high-water marks from the 1974 flood and from USGS topographic maps (1:24,000 scale) (Federal Emergency Management Agency, 1979, 1982). Field surveys and aerial photographs from near Interstate Highway 90 and upstream were used to develop a HEC-2 model and obtain cross sections. Bridges in the study reach also were surveyed. The FEMA model was calibrated using the high-water marks from the 1974 flood and other floods. Manning’s n roughness coefficients ranged from 0.025 to 0.035 for the main channel and 0.030 to 0.045 for the floodplain (Federal Emergency Management Agency, 1979, 1982). For the NF reach, coefficients of 0.040 were used for the main channel and ranged from 0.075 to 0.090 for the floodplain. For the SF reach, coefficients ranged from 0.035 to 0.065 for the main channel and 0.050 to 0.175 for the floodplain.

USGS Model (1992)

In 1992, the USGS developed a computer model of the lower Coeur d’Alene River using the FourPt program to estimate metal loading to Coeur d’Alene Lake (Woods and Beckwith, 1997). This model extended from the Cataldo to the Harrison gaging stations and comprised about 40 cross sections. The model also simulated the lateral lakes using cross sections. Manning’s roughness coefficients of 0.022 and 0.060 were used for the main channel and for the river banks, respectively. The model explicitly determined the discharge entering Coeur d’Alene Lake from the river at the Harrison gaging station. The Rose Lake gaging station provided river discharge data to estimate metal loading in the river and floodplain between the Rose Lake and Harrison gaging stations. Model boundaries used 15-minute discharge data from the Cataldo gaging station and stage data from the Harrison gaging station. However, the model was run using a 5-minute time step for model stability and convergence. The FourPt program (DeLong, and others, 1997) is an unsteady, 1D model for open channels based on the shallow surface-water equations. In backwater reaches, discharge is not a function of stage alone, and stage-discharge relations do not work. There may be many stages for one discharge or many discharges for one stage. FourPt can simulate flow conditions in various regular and irregular channels. The backwater in the river, caused by the lake and Post Falls Dam, prevents developing a unique stage-discharge relation at the Harrison and Rose Lake gaging stations. A model such as FourPt must be used to determine discharges at these backwater sites. No sediment transport was simulated, and the streambed in the model was unmovable (fixed bed). The model was used to determine discharge until 2005, when an ADVM was installed to monitor velocity at the Harrison gaging station.

University of Idaho Model (2004)

In 2004, the University of Idaho Center for Ecohydraulics Research used the MIKE11 model to develop a river model of the Coeur d’Alene River (Borden and others, 2004). This model extended from the Enaville gaging station on the NF Coeur d’Alene River and the Pinehurst gaging station on the SF Coeur d’Alene River to the Harrison gaging station on the main stem. The model used 161 cross sections. Cross sections were extended to include the floodplain by using data from USGS Digital Elevation Models. The lateral lakes also were included in the model as storage basins that required area-elevation tables for each lake. Manning’s roughness coefficients used ranged from 0.024 to 0.028 for the main channel and from 0.060 to 0.070 for the floodplain. The model used the 15-minute discharge data from the Enaville and Pinehurst gaging stations and 15-minute stage data from the Harrison gaging station as boundary conditions. The model also used the total sediment load curves from Clark and Woods (2001) at the upstream boundaries (Enaville and Pinehurst gaging stations).

The MIKE11 model simulates 1D unsteady flow in channels. It also can simulate erosion, deposition, and the transport of sediments in channels. Borden and others (2004) simulated two floods—a rain-on-snow flood (February 5–March 9, 1996) and a snowmelt flood (May 21–July 21, 1998) and produced longitudinal and planview animations of these floods. The flood from February 5 to March 9 extended across the valley, especially near Coeur d’Alene Lake.

Golder Associates Model (2005)

In 2005, Golder Associates, Inc. (2005) used the 1D steady-flow model, HEC-RAS, to model the lower Coeur d’Alene River to assess environmental effects of the Post Falls Dam. The purpose of this assessment was to support the license renewal of Avista Corporation’s hydroelectric facility at the Post Falls Dam on Coeur d’Alene Lake. The model used cross sections developed every 0.5 mi from river bathymetry data. Only in-channel flows were simulated by the model because ground elevation data for bank tops and floodplain were not available. A Manning’s n of 0.030 was assumed for cross sections at all flows. This model was not calibrated to hydraulic or sediment-transport conditions (Golder Associates, Inc., 2005). However, model simulation results were used to estimate flow velocities and to calculate sediment transport and incipient motion characteristics. Results showed that velocities in the river increased as lake levels decreased for a specified discharge, as is typical of backwater conditions. Results also indicated that Post Falls Dam did not significantly affect the transport of sediments in the Coeur d’Alene River.

Back to Table of Contents