U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2004–5237
Conceptual Model and Numerical Simulation of the Ground-Water-Flow System in the Unconsolidated Deposits of the Colville River Watershed, Stevens County, Washington
Summary
In recent years, increased use of ground- and surface-water supplies in watersheds of Washington State has created concern that insufficient instream flows remain for fish and other uses. Presently, surface water is available for further appropriation only from the mainstem of the Colville River from October 1 through July 15, all streams tributary to the Colville are considered to be fully appropriated by the Washington Department of Ecology under existing water rights, and issuance of new ground-water rights has been halted by the Washington Department of Ecology. The U.S. Geological Survey (USGS), in cooperation with the Colville River Watershed Planning Team, began a two part study in the summer of 2001 to investigate the ground-water system of the valley-fill deposits of the Colville River Watershed. Following the successful completion of the first phase, the USGS continued studies with the Colville River Watershed Planning Team to construct a steady-state, regional ground-water-flow model to develop a better understanding of the ground-water system and the potential regional effects of various ground-water use alternatives on the water resources of the Colville River Watershed.
The Colville River Watershed is underlain by unconsolidated deposits of glacial and non-glacial origin. Geologically, the basin can be grouped into three types of formations: bedrock, glacial deposits, and valley alluvium. The surficial geologic units and the deposits at depth were differentiated into aquifers and confining units on the basis of areal extent and general water-bearing characteristics. Five principal hydrogeologic units are recognized in the study area and form the basis of the ground-water-flow model: Upper outwash aquifer (UA), Till confining unit (TC), Colville Valley confining unit (VC), Lower aquifer (LA), and Bedrock (BR).
Ground-water flow in the unconsolidated sediments underlying the Colville River Watershed is simulated using the USGS modular three-dimensional finite-difference ground-water-flow model (MODFLOW). Five model layers were used to simulate the saturated unconsolidated sediments that overlie the bedrock and one layer of constant thickness was used to simulate the upper part of the bedrock. The boundary of the watershed was simulated as a no-flow boundary except in the area near the outlet where ground-water outflow is approximated. All major streams and lakes are included in the model as head-dependent flux boundaries. Ground-water recharge from precipitation was estimated using the USGS-developed Precipitation-Runoff Modeling System. Secondary recharge from septic systems, sewage disposal, and irrigation application was estimated from reported usage rates and from land-cover and land-use maps. Ground-water pumping from public-supply and domestic wells are included in the simulations. Initial hydraulic properties were estimated from specific capacity tests reported in the first phase of the Colville River Watershed study.
The steady-state Colville River Watershed ground-water-flow model was calibrated to September 2001 conditions using parameter estimation programs and methods that involve automated calibration procedures. Nonlinear regression analyses were applied to minimize the weighted differences, or residuals, between simulated and measured hydraulic head and streamflow measurements. Hydraulic-head measurements for calibration consisted of water-level measurements from 161 wells. To identify gaining and losing stream reaches, a low-flow seepage run (a set of streamflow measurements representing approximately steady-flow conditions) was conducted during September 2001. A total of 44 streamflow measurements were used in the calibration process.
Horizontal hydraulic conductivity values in the calibrated model ranged from 10 to 250 feet per day (ft/d) in the aquifer layers. Vertical hydraulic conductivity values of the confining units were estimated to be 0.0025 ft/d.
Simulated inflow to the model area was 53,000 acre-feet per year (acre-ft/yr) from precipitation and secondary recharge, and 36,000 acre-ft/yr from stream and lake leakage. Simulated outflow from the model was primarily through discharge to streams and lakes (71,000 acre-ft/yr), ground-water outflow (9,000 acre-ft/yr), and ground-water withdrawals (9,000 acre-ft/yr). The period of simulation, September 2001, was extremely dry so all components of the ground-water budget are presumably less than average flow conditions.
The challenge of the Colville River Watershed Planning Team is to supply additional water resources while limiting adverse effects on surface water and maintaining sufficient instream flows for all users. The numerical model was used to simulate the possible effects of increasing ground-water pumping. The model indicated that the increased pumping would come from reduced discharge to streams and lakes and reduced ground-water outflow. Because of the steady-state nature of the model, however, simulated effects are an "ultimate" response and do not indicate the time required to reach this result. The ultimate response simulated by the steady-state model constructed here – on the basis of conditions during a drier-than-average period?undoubtedly exceeds the actual response of the system to the simulated stress under average conditions.
The location of the simulated increase in ground-water pumping determined the primary source of the water withdrawn. Pumping wells in the northern end of the main Colville River valley derived a large proportion of the water from reduced ground-water outflow. Pumping wells located in the southern end of the main Colville River valley, however derived a large proportion of the water from reduced flow to rivers and streams. Alternatives simulating ground-water pumping in the relatively thin, discontinuous Upper aquifer resulted in the largest drawdowns.
The calibrated steady-state model was used to simulate ground-water-flow conditions as if no ground-water pumping, secondary recharge, or irrigation application occurred. In this simulation, the cumulative streamflow in the Colville River Watershed increased by 1.1 ft3/s, or about 37 percent of the present-day net ground-water withdrawal.
The model is intended to simulate the regional ground-water-flow system of the Colville River Watershed and can be used as a tool for water-resource managers to assess the regional effects of changes in stresses to the steady-state system. The regional scale of the model, coupled with relatively sparse data, spatially and temporally, must be considered when applying the model in less well understood areas or examining hydrologic conditions at a smaller scale than what is appropriate.
For more information about USGS activities in Washington, visit the USGS Washington District home page .
