USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2005-5227

Compilation of Information for Spokane Valley–Rathdrum Prairie Aquifer, Washington and Idaho

Back to Table of Contents

Hydrologic Information

Spokane Valley–Rathdrum Prairie Aquifer

The SVRP aquifer is composed of the previously described unconsolidated coarse-grained sand, gravel, cobbles, and boulders primarily deposited by a series of catastrophic glacial outburst floods. The material deposited in this high-energy depositional environment is coarser-grained than is typical for most basin-fill deposits, resulting in one of the most productive aquifers in the world. Water volume in the entire SVRP aquifer is estimated to be about 10 trillion gallons with an average of about 250 to 650 Mgal of water flowing through the aquifer daily at the Idaho–Washington border (MacInnis and others, 2000).

The aquifer generally has a greater percentage of finer material near the margins of the valley and becomes more coarse and bouldery near the center of the valley. Near Athol and between Athol and Rathdrum, several bodies of perched water have been reported in the flood deposits where downward water percolation is slowed by clay lenses or low-permeability till (Anderson, 1951). In the Spokane Valley, the aquifer is reported to contain no significant layers of low-permeability materials. In the Hillyard Trough, however, a clay layer appears to separate the aquifer into upper and lower units (pl. 2, section A-A’) (CH2M HILL, 2000). This clay layer is presumed to be glacial lake deposits from glacial Lake Columbia. In north Spokane County, the glaciolacustrine deposits are difficult to distinguish from the Latah sediments, especially in well logs (Boese, 1996).

In the Spokane area, CH2M HILL (1998) reports that the SVRP aquifer in the west part of the Spokane area consists of two relatively independent systems mostly separated by a buried basalt ridge that extends about 2 mi south of Five Mile Prairie. The main body of the aquifer is east of the ridge and in the Hillyard Trough and part of Spokane Valley. An area referred to as Trinity Trough is a breach across this basalt ridge and probably connects the east and west parts of the aquifer in that vicinity. The small portion of the aquifer between Spokane Falls and Nine Mile Dam has been referred to as to the western arm of the aquifer (J. Covert, Washington Department of Ecology, written commun., 2005) and the downriver segment (CH2M HILL, 1998).

In the downriver segment of the SVRP aquifer, at the former city of Spokane North Landfill site, monitoring well drillers’ logs indicate that SVRP aquifer flood deposits are underlain by glacial lake deposits (silt and clay) except in lower elevation areas near the Spokane River where the aquifer is underlain by basalt (CH2M HILL, 1988).

Areal Extent

The areal extent of the SVRP aquifer designations differ somewhat between investigators and over time. Many recent aquifer-related documents, including the SVRP Atlas (MacInnis and others, 2000, 2004), used a modified version of the original Spokane Valley–Rathdrum Prairie Sole Source Aquifer boundary designated by the USEPA in 1978 (fig. 5). Earlier studies tended to use a somewhat larger aquifer boundary as shown in Drost and Seitz (1978b, fig. 5). A slightly larger extent, that includes more of the surficial deposits adjacent to and in hydraulic connection with the Sole Source Aquifer, is shown in Berenbrock and others (1995, fig. 5).

For modeling purposes, it may be important to use a more inclusive aquifer boundary to better represent contributions from adjacent surficial deposits that are in hydraulic connection with the Sole Source Aquifer. An example of this is in the Chilco channel area where ground water flows into the Rathdrum Prairie portion of the Sole Source SVRP aquifer from an area in hydraulic connection with the aquifer but outside the official aquifer boundary (Painter, 1991a, 1991b, Graham and Buchanan, 1994). Similarly, Baldwin and Owsley (2005) have documented about 200 ft of saturated alluvium within the Middle (Ramsey) Channel that has similar water levels to the surrounding Sole Source Aquifer. They also suggested that this area be incorporated into the official aquifer extent in order for this area to be treated with the same levels of aquifer protection as the rest of the aquifer.

Figure 5. Boundary of the Sole Source Spokane Valley–Rathdrum Prairie aquifer, other study area extents, and generalized directions of ground-water flow, Washington and Idaho.

A revised extent map was drawn for this investigation that includes unconsolidated coarse-grained deposits based on the most recent surficial geologic mapping (pl. 2), including the Middle (Ramsey) and Chilco Channels. The revised map reflects adjustments that were made by moving the boundaries of the aquifer within Hoodoo and Spirit Valleys and near Careywood, Idaho, to ground-water divides that were mapped during previous investigations (Walker, 1964; Parliman and others, 1980).

The revised extent map also includes revisions to the aquifer boundary along the western arm of the aquifer near Riverside State Park in Washington (pl. 2). Recent analysis of ground-water level data, Nine Mile Reservoir elevation data, bedrock outcrops, and historical streamflow data have resulted in redrawing the western arm of the aquifer boundary ending near Nine Mile Falls rather than being continuous through that area as shown in previous aquifer boundaries. A bedrock ridge extending northwest from Five Mile Prairie to Nine Mile Dam forms the northern boundary of the western arm of the aquifer (J. Covert, Washington Department of Ecology, written commun., June 2005). The aquifer ends about one mile south of Nine Mile Dam and the water moving through this arm of the aquifer re-emerges into the Spokane River (Nine Mile Reservoir) before the dam. Summer streamflow data from the 1950s show that the Spokane River gained as much as 400 ft³/s between Hangman (Latah) Creek and Nine Mile Dam from the aquifer (J. Covert, Washington Department of Ecology, written commun., June 2005). Following completion of Nine Mile Dam, summer reservoir levels have been maintained at a constant level by changing the volume of water discharged through the dam. The water-table in the western arm of the aquifer is controlled by the elevation of Nine Mile Reservoir and ground water levels in the area between Seven Mile bridge and the dam follow the trend exhibited by Nine Mile Reservoir (J. Covert, Washington Department of Ecology, written commun., June 2005).

Aquifer Thickness

SVRP aquifer thickness generally is unknown except along its margins where wells have been drilled through its entire thickness. Since many wells in the SVRP aquifer are extremely productive, few wells in the aquifer extend more than 100 ft into the saturated zone. Well depths in the USGS National Water Information System (NWIS) database range from less than 10 to about 700 ft with a median of 162 ft. The greatest known thickness was recorded in Hillyard Trough where a 780-ft well did not penetrate the full thickness of the glacial and flood deposits (Cline, 1969). Coarse aquifer deposits in the Hillyard Trough area are separated by a layer of clay, silt, and sand at about 360 to 490 ft below land surface. The aquifer thickness is about 500 to 550 ft near the Idaho−Washington State line east of Spokane, Washington (Gerstel and Palmer, 1994) and overall, aquifer deposits are about 150-ft to more than 600-ft deep (MacInnis and others, 2000). Generalized hydrogeologic cross sections (simplified from CH2M HILL 1998, 2000a, and Baldwin and Owsley, 2005) for the Hillyard Trough (A−A’), Spokane Valley (B−B’), state line area (C−C’), and West (Main), Middle (Ramsey), and Chilco Channel areas (D−D’) are shown on plate 2.

To date (2004), the only aquifer-wide digital representation of the aquifer extent and bottom was prepared by Buchanan (2000), as input for a ground-water flow model of the SVRP aquifer system. His map of estimated bottom elevations of the aquifer’s bedrock base was based in part on seismic reflection profiling done in the 1990s (Buchanan, 2000). Although Buchanan’s work provides an aquifer-wide gross estimate of aquifer bottom, refinements are needed to better determine depth to bedrock, as well as determine the depth to the aquifer base where the aquifer is underlain by fine-grained unconsolidated material rather than consolidated bedrock.

In Washington, Poelstra and others (2005) have developed a preliminary, and as yet unreleased, three-dimensional digital geologic model for the Washington part of the SVRP aquifer system based on surface geology maps, geologic and geophysical cross sections, and well logs. Originally developed to evaluate the ground-motion amplification effect of soft soils in the upper 100 ft of the soil-rock column (Palmer and others, 2004), this three-dimensional model was expanded to include the full thickness of these “soft soils” as represented by unconsolidated deposits overlying bedrock. The preliminary model is composed of three separate unconsolidated units based on the dominant grain texture including gravel, sand, and silt-clay (S. Palmer, Washington Department of Natural Resources, written commun, February 2005). The gravel unit corresponds to SVRP’s major water productive section, as exhibited in the eastern part of the Spokane Valley. The sand unit consists of significantly thick slackwater flood deposits, typified by the thick accumulation exposed at the surface in the Hillyard Trough. The silt-clay unit corresponds to subsurface occurrences of Glacial Lake Columbia sediments present only in the subsurface in the Hillyard Trough and in the Little Spokane River drainage (S. Palmer, Washington Department of Natural Resources, written commun, February 2005).

Hydraulic Properties

A number of investigators have estimated hydraulic properties of the SVRP aquifer, including specific yield, hydraulic conductivity, and transmissivity. Hydraulic conductivity values for most of the central SVRP aquifer generally are large. Currently (2004), the USGS has information available in the NWIS database for about 1,200 wells inventoried as part of previous studies of the SVRP aquifer and surrounding basin-fill aquifers (table 3).

Table 3. Summary statistics for 1,190 wells completed in the Spokane Valley–Rathdrum Prairie aquifer and surrounding basin-fill aquifers.

[Summary statistics: From U.S. Geological Survey National Water Information System database. Abbreviation: <, less than]

Wells in the aquifer generally yield large volumes of water with relatively little drawdown. Wells in the saturated coarse-grained deposits can yield several thousand gallons per minute (Bolke and Vaccaro, 1979; Stone and others, 1996), with several wells near Spokane reportedly yielding over 5,000 gal/min. Reported yields for wells in the SVRP aquifer in the USGS database range from less than 2 to 25,000 gal/min with a median of 500 gal/min (table 3).

Although much of the SVRP aquifer is considered highly permeable, hydraulic characteristics are locally variable and include less-permeable, fine-grained sedimentary material. Several previous studies including Drost and Seitz (1978b), Bolke and Vaccaro (1981), and CH2M HILL (1998) have calculated aquifer characteristics based on aquifer tests and ground-water model simulations. Although hydraulic properties of the aquifer were variable, most results indicated that hydraulic conductivity (a measure of the ability of the aquifer material to transmit water) and transmissivity (the rate at which water is transmitted through a unit width of the aquifer under a unit hydraulic gradient, equal to the hydraulic conductivity multiplied by the aquifer thickness) values were on the upper end of values measured in the natural environment. Drost and Seitz (1978b) reported transmissivity values that ranged from less than 130,000 ft²/d in the western part of the aquifer to more than 13 million ft²/d near the Washington–Idaho state line. Estimated ground-water velocities exceeded 60 ft/d near the state line to about 47 ft/d in the Hillyard Trough (fig. 1). Bolke and Vaccaro (1981) estimated hydraulic conductivity values of between about 2,600 to 6,000 ft/d for most of the aquifer on the Washington side, with a value of about 860 ft/d in the less permeable Hillyard Trough area. CH2M HILL (1998) reported hydraulic conductivity values ranging from about 100 to 6,200 ft/d, with most values greater than 1,000 ft/d.

Vertical anisotropy is the ratio of horizontal hydraulic conductivity to vertical hydraulic conductivity. Bolke and Vaccaro (1981) stated the available data suggested no vertical stratification of the aquifer lithology, and therefore, no vertical anisotropy. Golder Associates, Inc. (2004) used an initial vertical anisotropy of 3:1. CH2M HILL (1998) assumed a vertical anisotropy of 10:1 that produced conservative (large) estimates of well capture zones.

Ground-Water Occurrence and Movement

Depth to ground water in the SVRP aquifer ranges from near land surface to more than 500 ft below land surface (Bolke and Vaccaro, 1979; Berenbrock and others, 1995; Briar and others, 1996; Stone and others, 1996; MacInnis and others, 2000). The greatest depth to ground water occurs in the northern Rathdrum Prairie in Idaho and the shallowest depth (less than 50 ft in places) is near Spokane along the Spokane River. Water levels measured in wells and recorded in the USGS database (2004) range from 2- to 573-ft deep with a median depth of 106 ft (table 3). Seasonal water-level fluctuations in the aquifer generally are less than 15 ft in most areas (Drost and Seitz, 1978b).

The water table in the SVRP aquifer generally reflects the land-surface topography and slopes from Hoodoo Valley and Lake Pend Oreille, Idaho, to Nine Mile Falls, Washington. Ground water generally flows in a southward direction from the area near the southern end of Lake Pend Oreille with a water-level altitude of about 2,150 ft, towards the city of Coeur d’Alene, and then westward towards the city of Spokane with an altitude of about 1,500 ft near the Little Spokane River (Drost and Seitz, 1978b; Molenaar, 1988; and MacInnis and others, 2000). The water table in the northeastern-most part of the aquifer slopes about 20 ft/mi, while the major part, from north of Round Mountain, Idaho, to the southern end of the Hillyard Trough, Washington, slopes gently from 2 to 10 ft/mi (Drost and Seitz, 1978b). Steeper slopes, sometimes more than 60 ft/mi, are in the Hillyard and Trinity Trough areas and along the Spokane River west of Five Mile Prairie. Generalized ground-water flow directions based on water-level elevations measured in the aquifer are shown in figure 5.

SVRP aquifer upstream margins include the southern parts of the Hoodoo and Spirit Valleys and Careywood in southern Bonner County, Idaho, in addition to the outlet at the southern end of Lake Pend Oreille (pl. 2). In the Cocolalla Valley, a ground-water divide is reported near Careywood along U.S. Highway 95 (Parliman and others, 1980). Ground water north of the divide flows northeast toward the Pend Oreille River; ground water south of the divide flows southwest toward Athol and the main body of the SVRP aquifer. In the Hoodoo Valley, historical water-level elevations indicated that a water-table divide was between Edgemere and Harlem (Walker, 1964). Ground water north of the divide moved northward toward the Pend Oreille River; ground water south of the divide moved southward toward Athol. In Spirit Valley, the ground-water divide was near Blanchard Lake (Parliman and others, 1980). West of the divide, ground water flows northwestward toward the Pend Oreille River; east of the divide, ground water flows southeastward into the main body of the SVRP aquifer.

Hammond (1974) reported that ground-water flow can be variable between Lake Pend Oreille and Athol. When the lake level declines below the adjacent water table an apparent ground-water mound (divide) is formed near Farragut State Park (Hammond, 1974). At such times, ground water can flow toward the lake from a distance of at least one-third of a mile and possibly farther (Hammond, 1974).

Most ground water from the SVRP aquifer discharges either to the Spokane and Little Spokane Rivers or is withdrawn by wells (Drost and Seitz, 1978b; Molenaar, 1988). An unknown amount of ground water may leave the system from the lower part of the aquifer in the Hillyard Trough and Little Spokane River area near Long Lake. Recharge occurs by infiltration of precipitation, snowmelt, irrigation water, subsurface inflow from adjoining highlands and tributary valleys, and leakage from adjacent and overlying surface-water sources (Molenaar, 1988). Several investigations reported that the Spokane River varies from losing to gaining along its course as it flows over the SVRP aquifer (Gearhart and Buchanan, 2000; Marti and Garrigues, 2001; and Caldwell and Bowers, 2003).

Ground-Water Levels

In 2004, an observation network was established by the USGS to measure water levels within the SVRP aquifer and adjacent hydrogeologic units. A monthly network of 47 wells was established in June 2004 to manually measure depth to water in wells throughout the aquifer. In July 2004, an eight-well network was established where depth to water is recorded hourly. As of 2005, both monthly and recorder networks are ongoing. During one week in September 2004, the depth to water was measured in 268 wells to develop a water-table map of the SVRP aquifer. Wells were visited by personnel from IDWR, WADOE, and USGS. The well locations in the observation networks and wells measured in September 2004 are shown on plate 1. The September 2004 water-table map and the supporting data are reported in Campbell (2005).

In addition to water-level data collected by USGS, water-level data are available from other sources for hundreds of wells throughout the SVRP aquifer. These data are available from several local and State government agencies, colleges and universities, water purveyors, and environmental consulting firms. Wells with multiple water-level measurements and measurement frequencies ranging from every 15 minutes (continuous recorders) to every few months are available for more than 100 wells. City and county government agencies have collected water levels as part of several programs including landfill monitoring, ground-water/ surface-water interaction studies, and wellhead protection. Water-level data from statewide monitoring well networks and local investigations are available from State agencies like IDWR and WADOE. Universities and colleges have collected water-level data as part of graduate thesis studies, class projects, and projects conducted by faculty members. Several water purveyors have been proactive in monitoring water levels in public water-supply and monitoring wells: some purveyors have collected over twenty years of data. Environmental consulting firms have collected water-level data from parts of the SVRP aquifer for several studies

Lake Levels

Lake stage data collection has been continuous at Coeur d’Alene, Hayden, and Pend Oreille Lakes, Idaho, and Long Lake, Washington, since the early 1900s (table 4, fig. 6). Historical data are available for Twin Lakes, Idaho, and Newman and Liberty Lakes, Washington, from the 1950s to the late 1960s or 1980s (table 4, fig. 6). As part of this study, gaging stations were re-established at Liberty and Newman Lakes, Washington, in August 2004, and at Twin Lakes, Idaho, in October, 2004. Also, in October 2004, gaging stations were established at Spirit and Hauser Lakes, Idaho, where no gaging stations were located previously. Currently (2005), USGS is making monthly lake stage measurements at Spirit, Twin, Hayden, and Hauser Lakes, Idaho, and Newman and Liberty Lakes, Washington. Hourly measurements are recorded at long-term measurement sites at Lakes Coeur d’Alene (USGS) and Pend Oreille (U.S. Army Corps of Engineers), Idaho and Long Lake (Avista Corporation), Washington. Liberty Lake Sewer and Water District began monitoring lake stage of Liberty Lake, Washington, January 1, 2004 (Liberty Lake Sewer and Water District, accessed June 1, 2005, http://207.88.115.227/libertylakemonitoring/liberty_lake.htm).

Table 4. Summary of historical or project-related lake stage data collection sites for lakes along the margins of the Spokane Valley–Rathdrum Prairie aquifer, Washington and Idaho.

[Locations of stations are shown on figure 6. Period of record: At the time of publication of this report, the period of record is expected to continue beyond 2005 at each of these sites. Abbreviations: ID, Idaho; WA, Washington; USGS, U.S. Geological Survey]

Figure 6. Locations of long-term and project-related lake-stage measurement sites, Washington and Idaho.

Aquifer Boundaries

In most places, the Spokane Valley–Rathdrum Prairie aquifer is bounded laterally by bedrock and the lower or bottom aquifer boundary is mostly unknown except along the margins or in shallower parts of the aquifer where wells have penetrated the entire aquifer thickness and reached bedrock or silt and clay deposits. Reported ground-water divides approximately represent the aquifer boundary in the Hoodoo and Spirit Valleys and near Careywood, Idaho (Walker, 1964 and Parliman and others, 1980). Upgradient aquifer areas also are bounded by tributary lakes, including Pend Oreille, Spirit, Twin, Hayden, Coeur d’Alene, Hauser, Newman, and Liberty. Streams tributary to the aquifer include Lewellen, Sage, and Rathdrum Creeks in Idaho, and Chester and Saltese Creeks in Washington. Streams tributary to the Spokane River in the aquifer extent include Hangman (Latah) Creek near Spokane, Washington, and the Little Spokane River north of Spokane. The aquifer’s lower discharge area is near Long Lake at the confluence of the Spokane and Little Spokane Rivers.

Water-Budget Components

A water budget is an accounting of water and its movement in a hydrologic system. Water budgets can be as simple as a few numbers representing water added to and subtracted from the hydrologic system, or as complex as a numerical simulation of the hydrologic system. This hydrologic system can range in scale from global to site specific and could include only ground water, only surface water, or both. A water budget is a useful tool for helping water-resource scientists and managers conceptualize the hydrologic system. Because some of the inflows and outflows from the system cannot be measured directly, however, they must be estimated. Therefore, the resulting water budget is an approximation of the physical hydrologic system, and the measured inflow and outflow totals may not balance exactly.

Several investigators have compiled complete water budgets for large parts of the SVRP aquifer (table 5). These water budgets are not directly comparable because most studies have used different boundaries for the SVRP aquifer. More than 25 publications have addressed aspects of the water budget for the SVRP aquifer using different techniques. Water-budget component estimates from previous studies are summarized in tables 6 through 13.

Because of the long history of water development in the study area and the obscure nature of many reports on water-resource issues, numerous discrepancies exist between various reports. These discrepancies likely are due to the unavailability of original reports, resulting in investigators citing data referenced in subsequent reports. This potential source of error, compounded with misunderstandings and typographic mistakes, likely caused reported values to “drift” from the original. For this report, original sources were reviewed in nearly all cases; however, some documents were unavailable. These references are identified in the tables 6 through 13. How a particular value was obtained is not always known, but where possible, methods used by various investigators for their estimates have been identified. Finally, different investigators have defined or grouped measurement areas differently and estimates should be compared with caution. This report is intended as a review—the interested reader is urged to check original sources.

Recharge and Underflow—How Ground Water Enters the Aquifer

Water enters an aquifer by many processes and settings and these processes and settings can be classified in numerous ways. In this report, recharge is discussed by the setting in which it occurs. Three main settings in which water enters the SVRP aquifer are: the valley floor over the aquifer, tributary basins, and adjacent uplands surrounding the aquifer, and the Spokane River itself (which also acts as a discharge setting in some reaches). Some investigators have included total recharge estimates to the aquifer without differentiating the source. These estimates are included in table 5.

Table 5. Published water budgets for the Spokane Valley–Rathdrum Prairie aquifer.

[Method: M, ground-water-flow model; R, referenced; SM, streamflow measurements; W, water balance; WY, watershed yield. Abbreviations: ft³/s, cubic foot per second; –, not applicable or unknown]

¹ Recharge only.

² Does not include Lake Pend Oreille.

³ From Frink (1964).

⁴ Steady state.

⁵ Transient.

Valley Floor

The main source of recharge to ground water from the valley floor is infiltration from land surface: precipitation, applied irrigation water, canal-seepage loss, and septic-tank effluent. However, water loss to evapotranspiration decreases the amount of precipitation and irrigation water that reaches the saturated zone.

Precipitation

Direct recharge from precipitation on the valley floor was recognized early as a significant component of SVRP water budgets due to the scarcity of streams that reach the Spokane River. Recharge estimates from precipitation are more straightforward than many other water-budget components; however, previous estimates still range over an order of magnitude (table 6), partly due to differences in how areas are delineated—some investigators included the entire watershed while others included only the valley floor. Also, some investigators factored evapotranspiration losses directly into their recharge estimates while others subtracted evapotranspiration as a separate item.

A facet of precipitation recharge is the presence of numerous storm-water injection, or dry wells in the study area. These drain wells allow local disposal of storm water eliminating the need for extensive storm-water sewerage. Due to the long-standing use of these dry wells, it is unclear exactly how many exist, though McLeod (1991) reported about 2,500 in the “Panhandle area around Coeur d’Alene.” Golder Associates, Inc. (2004) used Spokane County GIS coverages for the location of these wells and calculated that they recharge 83-87 percent of precipitation in the capture zone of the well and modeled them explicitly with a density of as many as 87 wells per square mile.

Applied Irrigation Water and Canal Seepage

Many early studies of the SVRP aquifer were related to irrigation projects, consequently several estimates have been made of irrigation seepage and canal leakage. These irrigation-related components are summarized in table 6.

Septic-System Effluent

Recharge from septic-system effluent was not included explicitly as a component in SVRP aquifer water budgets until around the advent of numerical simulation. This probably corresponds to rapid population growth and explains why previous investigators may not have considered it significant. Septic-system effluent recharge estimates are shown in table 6.

Inflow from Tributary Basins, Adjacent Uplands, and Subsurface

Previous SVRP aquifer water budgets indicate that recharge from lakes and streams in tributary basins and adjacent uplands are the largest source of water to the SVRP aquifer. Traditionally, the quantity of such recharge has been difficult to measure and often has been calculated indirectly through water-budget or modeling methods.

Because of the SVRP aquifer’s physical setting, many investigators have believed that Coeur d’Alene Lake and Lake Pend Oreille are among the largest sources of recharge to the aquifer. Consequently, much effort has gone into quantifying recharge contribution by these and other lakes using methods ranging from simple watershed-yield estimates to numerical simulation (table 7). Note that some estimates for a given lake may range over several orders of magnitude.

Tributary basins without major lakes and adjacent uplands, while individually small, collectively contribute a significant quantity of recharge to the SVRP aquifer (table 8). However, they have received less attention than other water-budget components possibly because quantifying such recharge is difficult or because study areas differ. Previous seepage estimate studies primarily used watershed-yield, streamflow measurement, or numerical simulation methods and are shown in table 8. Some estimates range over an order of magnitude. Of special interest is possible underflow from the Hoodoo Valley into the SVRP aquifer. Though surface-water drainage is to the north, several investigators suggested that ground water flows south, aided by the increased head of water behind Albeni Falls dam.

Underflow into the SVRP aquifer cannot be directly measured; therefore, these estimates tend to be the residual of other water-budget components in ground-water-flow models. Table 9 shows underflow recharge estimates by previous investigators.

Spokane River

The direction and amount of water flowing between the Spokane River and the SVRP aquifer is one of the most important hydrologic issues in the study area. Not only do the volume and direction of flow between surface and ground water affect the amount of water in the river, they affect the volume of ground water available in the aquifer. Numerous studies have examined this interaction with a variety of techniques over different river reaches and have defined gaining (the river gains flow from ground-water discharge) and losing (the river loses flow to ground-water recharge). These often conflicting numbers result from different reach definition, seasonal and yearly precipitation variance, development in the study area, and study method. For instance, most investigators for most time periods have concluded that the Spokane River between Post Falls and Nine Mile Falls gains water from the aquifer, though shorter reaches lose water to the aquifer. Various estimates by previous investigators are shown in table 10 and are discussed in more detail in the Ground-Water/Surface-Water Interactions section.

Table 6. Estimates of total recharge, land-surface recharge, and evapotranspiration for the Spokane Valley–Rathdrum Prairie aquifer.

[Recharge or discharge: Negative values indicate discharge from the aquifer. Method: C, calculated; D, Darcy’s Law; M, ground-water-flow model; MO, meteorological observations; R, referenced; SM, streamflow measurements; W, water balance; WY, watershed yield. Abbreviations: WRIA, Water Resource Inventory Area; ID, Idaho; WA, Washington; ft³/s, cubic foot per second; –, not applicable or unknown]

¹ Steady-state.

² Transient.

³ Range of monthly averages.

⁴ From Frink (1964).

⁵ From Anderson (1951).

Table 7. Estimates of seepage from lakes into the Spokane Valley–Rathdrum Prairie aquifer.

[Method: D, Darcy’s Law; M, ground-water-flow model; R, referenced; W, water balance; WY, watershed yield. Abbreviations: ft³/s, cubic foot per second; ≥, greater than or equal; ≤, less than or equal; –, not applicable or unknown]

¹ From Frink (1964).

² From Anderson (1951).

³ From Hammond (1974).

⁴ From Painter (1991).

Table 8. Estimates of recharge from tributary streams, drainages, and uplands into the Spokane Valley–Rathdrum Prairie aquifer.

[Method: M, ground-water-flow model; R, referenced; SM, streamflow measurements; WY, watershed yield. Abbreviations: ft³/s, cubic foot per second; –, not applicable or unknown]

Table 9. Estimates of non-tributary underflow into and out of the Spokane Valley–Rathdrum Prairie aquifer.

[Recharge: Negative values indicate discharge from the aquifer. Method: M, ground-water-flow model; R, referenced. Abbreviation: ft³/s cubic foot per second]

¹ Steady-state.

² Transient.

Table 10. Estimates of seepage between the Spokane River and the Spokane Valley–Rathdrum Prairie aquifer.

[Recharge: Negative values indicate discharge from the aquifer. Method: D, Darcy’s Law; M, ground-water-flow model; CM, chemical mass balance; R, referenced; SM, streamflow measurements. Abbreviations: RM, river mile; SIRTI, Spokane Intercollegiate Research and Technology Institute; WWTP, wastewater treatment plant; ft³/s, cubic foot per second; ≤, less than or equal; –, not applicable or unknown]