INTRODUCTION

The effects of ground-water withdrawals on streamflow have become an important issue in the Puget Sound Lowland of western Washington. Surface-water resources are fully allocated in many parts of the region where population growth has increased the demand for water, and future growth will most likely depend on the availability of ground water. The purpose of this study was to gain a better understanding of the relations and interactions between ground-water and surface-water systems in small basins of the Puget Sound Lowland. It was also hoped that this study, conducted in cooperation with the Washington State Department of Ecology (Ecology), would identify some of the important factors controlling the response of the systems to ground-water withdrawals.

Background

In western Washington, as in many areas of the United States, water users, developers, and regulators are confronting questions about the effects of ground-water withdrawals on ground-water levels and streamflow. Regulators, such as Ecology, are charged with the responsibility of limiting these effects to acceptable levels. This task generally requires the regulators to analyze the effects of proposed ground-water withdrawals on streamflow and spring discharge. The site-specific data on geology, ground water, and streamflow to support these analyses are usually not readily available and can be costly to collect. Over the years, many methods have been devised to estimate the response of surface-water systems to ground-water withdrawals, ranging from relatively simple analytical methods, such as the one advanced by Jenkins (1970), to site-specific transient three-dimensional numerical models. The drawbacks of these methods are that they are either too simplistic to be applied in the complex hydrogeologic environments found in the Puget Sound Lowland, or that they are too costly, time consuming, and their results are not transferable between basins (for example, site-specific models). The difficulty in finding a suitable means of estimating ground water-surface water interactions stems in part from the irregular nature of the quasi-layered glacial deposits. The complex assemblage of these deposits makes numerical simulations, analytical solutions, and intuitive assessments difficult to apply and interpret. The inherent difficulty in assimilating the many factors involved in ground water-surface water interactions makes the issue of ground-water rights versus surface-water rights one of the most intractable problems facing water-supply managers and regulators.

A large part of the flow of streams originating in the Puget Sound Lowland consists of ground water discharged from aquifers of unconsolidated Pleistocene glacial outwash deposits. This part of streamflow is termed baseflow. The water in these streams is used for drinking water, irrigation, and industry, and is appropriated (legally "set aside") for water users through a permit system administered by Ecology. Withdrawals from many streams are limited by State regulations that prohibit users from withdrawing water when the stream has receded to a prescribed minimum acceptable flow. The minimum flows were established so that enough water remains in the stream to allow for the passage of anadromous fish (for example, salmon), the dilution of wastes, and other instream uses. In most cases, the total amount of water that has been appropriated from a stream exceeds the amount available (the amount in excess of the minimum acceptable flow) during periods of low flow, and many of these streams have been closed to further appropriation. Nevertheless, the population continues to increase in the Puget Sound Lowland, and in areas where streamflow is no longer available, water managers, developers, and individuals in need of new water supplies are requesting ground-water-withdrawal permits from the State. There is concern that development of ground water as a water supply may lower ground-water levels and consequently the baseflow of streams in some basins. This would reduce the availability of surface water to existing users and could reduce baseflows to levels below the established minimum flow during some periods. In order to allow development of ground-water resources while ensuring acceptable amounts of baseflow in regulated streams, Ecology needs to estimate the potential for a proposed withdrawal to reduce baseflows.

Purpose and Scope

The purpose of this report is to describe the results of a study to improve the fundamental understanding of ground-water flow and the effects of ground-water withdrawals on ground-water discharge to streams and springs in small Puget Sound Lowland basins. The specific objectives of this study were (1) to develop a generalized conceptual model of the hydrogeology of basins in the Puget Sound Lowland; and (2) to demonstrate the effects of various ground-water withdrawal scenarios on the baseflow of streams originating in these basins---specifically, to evaluate the effects of variations in well location, depth of completion, and rate of withdrawal on the rate and distribution of ground water discharged as baseflow. These evaluations will help improve the understanding of ground water-surface water relations in basins with glacial geology typical of western Washington, and provide regulators a means of assessing the soundness and usefulness of existing or proposed regulation or permitting schemes.

Description of the Puget Sound Lowland

The Puget Sound Lowland, as defined by Vaccaro (J.J. Vaccaro, U.S. Geological Survey, written commun., 1993), is an elongate basin that extends approximately 200 mi along its north-south axis from the Fraser River in Canada to the southern extent of Pleistocene glaciation near Centralia, Wash. (fig. 1). The area shown on figure 1 extends from the crest of the Cascade Range on the east and to the Olympic Mountains and the Straits of Juan De Fuca and Georgia on the west, and covers about 17,600 mi2. The part of the Puget Sound Lowland underlain by Quaternary glacial sediments, which make up the principal aquifers, was the focus of this study. This part of the Puget Sound Lowland ranges in width from about 15 to 80 mi and covers an area of nearly 7,200 mi2.

During the Pleistocene epoch, southward moving continental glaciers covered the lowland numerous times. Most aquifers and many of the confining layers in the lowland are composed of unconsolidated sedimentary materials deposited as a result of the glaciers' passage. The depositional processes associated with the glaciers produced the layering which is characteristic in the lowland. Periods when the glaciers were advancing or retreating are associated with layered deposits of sand or gravel and till, and periods when glaciers were not in the area, or far removed from it, are associated with fine-grained lacustrine deposits.

The topography of the lowland has been shaped by deposition and erosion that has occurred during the 12,000 to 13,000 years since the last glaciation. The lowland is generally characterized by flat, featureless drift plains that lie at altitudes of 200 to 600 ft above sea level. In places, the drift plains have been incised by major stream valleys; steep bluffs form the boundaries between the drift plains and the major stream valleys below. The effects of continental glaciation on the topography of the lowland are evident in the predominant north-south and northwest-southeast alignment of lakes, ridges, and major stream valleys that were etched by moving ice. As they cross the drift plains, streams have low hydraulic gradient, but the gradient steepens as the streams descend from the plains to the major stream valleys below.

The Puget Sound Lowland has a mid-latitude, Pacific-coast-marine type climate characterized by warm, dry summers and cool, wet winters. Mean annual precipitation ranges from about 25 to 60 in/yr, with a mean of 38 in/yr in 26 drainage basins within the Puget Sound Lowland (J.J. Vaccaro, U.S. Geological Survey, written commun., 1993). Nearly 80 percent of annual precipitation falls between October and March. Summer temperatures range from 60o F to 80o F, and winter temperatures range from 30o F to 50oF.

Where soils are poorly drained, native vegetation includes fir, cedar, alder, and madrona with an understory of huckleberry, Oregon grape, salal, and blackberry. On well-drained soils underlain by coarse-grained outwash deposits, the dominant vegetation consists of wild grasses, bracken fern, and scotch broom with patches of fir and oak.

In 1990, water use in the Puget Sound Lowland was 810 Mgal/d (J.J. Vaccaro, U.S. Geological Survey, written commun., 1993), 21 percent (174 Mgal/d) of which was ground water supplied by public water systems and 22 percent (178 Mgal/d) of which was ground water from private water systems. The total ground-water withdrawal of 352 Mgal/d in 1990 was approximately three times the amount supplied from ground-water sources in 1965.

Method of Study

Following an extensive review of the literature on the hydrogeology of the lowland, a conceptual model of a small basin was developed that incorporated all of the most significant hydrogeologic characteristics. These characteristics included the glacial stratigraphy unique to the region, the drift plain-bluff-valley topography created by the glaciation and subsequent fluvial erosion and deposition, and the stream networks that provide the surface-water and ground-water drainage for the systems.

References to the many studies of the geology and hydrology of the Puget Sound Lowland can be found in a bibliography compiled by Jones (1991) as part of a regional aquifer system analysis (RASA) carried out by the U.S. Geological Survey. Vaccaro (J.J. Vaccaro, U.S. Geological Survey, written commun., 1993) summarizes many of the results and conclusions from the RASA study, including analyses using cross-sectional numerical models in various hydrogeologic settings in the Puget Sound Lowland. This work provided much of the basis for the conceptual model used in the study. The findings of Dion and others (1994) for northern Thurston County and Woodward and others (1995) for southwestern King County also greatly influenced the conceptual model.

The criteria used in developing the conceptual model of the basin were (1) the basin had to be defined in sufficient detail to incorporate the salient features that control ground-water flow, and (2) the definition had to be general enough to be representative of a typical lowland basin. Attaining a balance between simplicity and detail in the conceptual and numerical models was key to producing useful results from the model analysis. Whereas detail was required to provide realistic boundary conditions, simplicity was essential for interpretation of cause and effect relations from model results.

A scale was chosen for the basin that would allow analysis of ground-water development scenarios ranging from single-well, local-scale withdrawals to multiple-well, basin-scale withdrawals. The physical attributes of the hypothetical basin, including topography, geology, and drainage, were then synthesized on the basis of the conceptual model. The conceptual model provided the guidelines such as the altitude and slope of the land surface, stream gradients and tortuosity, and thickness and extent of geologic layers. The spatial data describing the basin were compiled, checked, and stored in digital form using a geographic information system (GIS); this system was later used to create the data files needed by the numerical ground-water model and to store, display, and analyze the results of the model.

Initial estimates of recharge, hydraulic characteristics, and boundary conditions used in the model were based on typical values found by previous investigators in the Puget Sound Lowland. Most of these values were modified during calibration to make the hydrologic conditions simulated by the model more closely match conditions found in small Puget Sound Lowland basins; modifications were generally minor and always left the model parameters well within the range that would be expected in the Puget Sound Lowland for similar conditions or materials. The model was calibrated for predevelopment, steady-state hydrologic conditions. The parameter-adjustment process was completed when the simulated conditions matched expected conditions within tolerable limits.

The hydraulic heads and discharges to streams and springs simulated by the predevelopment, steady-state model represented the baseline hydrologic conditions used in the analysis of the hydrologic response of the hypothetical basin to ground-water development. Seven series of simulations were designed to analyze the effects of specific development variables, such as well depth or distance from a stream, on the response of the ground-water system. The response of the system to each scenario was evaluated by comparing simulated heads and ground-water discharge to streams and springs with those from the baseline model. The location and magnitude of reductions of natural ground-water discharge to streams and springs and the extent and magnitude of water-level declines were compared for each series of simulations. A total of 30 simulations were made in 7 series.

Acknowledgments

Discussions with Linton Wildrick, Robert Garrigues, and Kirk Sinclair of the Washington State Department of Ecology were helpful in development of the scenarios simulated with the numerical model and in determining the best methods of presenting the results of the simulations. Exchanges with members of the Technical Advisory Committee for Hydraulic Continuity, Washington State Water Resources Forum, provided insight into the nature and importance of ground water-surface water issues in the Puget Sound Lowland.