ABSTRACT

The upper Deschutes Basin is among the fastest growing regions in Oregon. The rapid population growth has been accompanied by increased demand for water. Surface streams, however, have been administratively closed to additional appropriation for many years, and surface water is not generally available to support new development. Consequently, ground water is being relied upon to satisfy the growth in water demand. Oregon water law requires that the potential effects of ground-water development on streamflow be evaluated when considering applications for new ground-water rights. Prior to this study, hydrologic understanding has been insufficient to quantitatively evaluate the connection between ground water and streamflow, and the behavior of the regional ground-water flow system in general. This report describes the results of a hydrologic investigation undertaken to provide that understanding. The investigation encompasses about 4,500 square miles of the upper Deschutes River drainage basin.

A large proportion of the precipitation in the upper Deschutes Basin falls in the Cascade Range, making it the principal ground-water recharge area for the basin. Water-balance calculations indicate that the average annual rate of ground-water recharge from precipitation is about 3,500 ft³/s (cubic feet per second). Water-budget calculations indicate that in addition to recharge from precipitation, water enters the ground-water system through interbasin flow. Approximately 800 ft³/s flows into the Metolius River drainage from the west and about 50 ft³/s flows into the southeastern part of the study area from the Fort Rock Basin. East of the Cascade Range, there is little or no ground-water recharge from precipitation, but leaking irrigation canals are a significant source of artificial recharge north of Bend. The average annual rate of canal leakage during 1994 was estimated to be about 490 ft³/s. Ground water flows from the Cascade Range through permeable volcanic rocks eastward out into the basin and then generally northward. About one-half the ground water flowing from the Cascade Range discharges to spring-fed streams along the margins of the range, including the upper Metolius River and its tributaries. The remaining ground water flows through the subsurface, primarily through rocks of the Deschutes Formation, and eventually discharges to streams near the confluence of the Deschutes, Crooked, and Metolius Rivers. Substantial ground-water discharge occurs along the lower 2 miles of Squaw Creek, the Deschutes River between Lower Bridge and Pelton Dam, the lower Crooked River between Osborne Canyon and the mouth, and in Lake Billy Chinook (a reservoir that inundates the confluence of the Deschutes, Crooked, and Metolius Rivers).

The large amount of ground-water discharge in the confluence area is primarily caused by geologic factors. North (downstream) of the confluence area, the upper Deschutes Basin is transected by a broad region of low-permeability rock of the John Day Formation. The Deschutes River flows north across the low-permeability region, but the permeable Deschutes Formation, through which most of the regional ground water flows, ends against this rampart of low-permeability rock. The northward-flowing ground water discharges to the streams in this area because the permeable strata through which it flows terminate, forcing the water to discharge to the surface.Virtually all of the regional ground water in the upper Deschutes Basin discharges to surface streams south of the area where the Deschutes River enters this low-permeability terrane, at roughly the location of Pelton Dam.

The effects of ground-water withdrawal on streamflow cannot presently be measured because of measurement error and the large amount of natural variability in ground-water discharge. The summer streamflow near Madras, which is made up largely of ground-water discharge, is approximately 4,000 ft³/s. Estimated consumptive ground-water use in the basin is about 30 ft³/s, which is well within the range of the expected streamflow measurement error. The natural variation in ground-water discharge upstream of Madras due to climate cycles is on the order of 1,000 ft³/s. This amount of natural variation masks the effects of present ground-water use. Even though the effects of ground-water use on streamflow cannot be measured, geologic and hydrologic analysis indicate that they are present.

Ground-water-level fluctuations in the upper Deschutes Basin are driven primarily by decadal climate cycles. Decadal water-level fluctuations exceeding 20 ft (feet) have been observed in wells at widespread locations near the margin of the Cascade Range. The magnitude of these fluctuations diminishes toward the east, with increasing distance from the Cascade Range. Annual water-level fluctuations of a few feet are common in areas of leaking irrigation canals, with larger fluctuations observed in some wells very close to canals. Annual water-level fluctuations of up to 3 ft due to ground-water pumping were observed locally. No long-term water-level declines attributable to pumping were found in the upper Deschutes Basin.

The effects of stresses to the ground-water system are diffused and attenuated with distance. This phenomenon is shown by the regional response to the end of a prolonged drought and the shift to wetter-than-normal conditions starting in 1996. Ground-water levels in the Cascade Range, the locus of ground-water recharge, stopped declining and started rising during the winter of 1996. In contrast, water levels in the Redmond area, 30 miles east of the Cascade Range, did not start to rise again until late 1997 or 1998. The full effects of stresses to the ground-water system, including pumping, may take several years to propagate across the basin. Ground-water discharge fluctuations were analyzed using stream-gage records. Ground-water discharge from springs and seeps estimated from stream-gage records shows climate-driven decadal fluctuations following the same pattern as the water-level fluctuations. Data from 1962 to 1997 show decadal-scale variations of 22 to 74 percent in ground-water discharge along major streams that have more than 100 ft³/s of ground-water inflow.