Scientific Investigations Report 2007–5050
U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2007–5050
Version 1.1, April 2010
Back to Table of Contents
The principal geologic factors that influence ground water are the porosity and permeability of the rock or sediment through which it flows. Porosity, in general terms, is the proportion of a rock or deposit that consists of open space. In a gravel deposit, this would be the proportion of the volume of the deposit represented by the space between the individual pebbles and cobbles. Permeability is a term used to describe the ease with which a fluid can move through a material such as rock or sedimentary deposits. Deposits with large interconnected open spaces, such as gravel, offer little resistance to ground-water flow and are, therefore, usually highly permeable. Rocks with few, very small, or poorly connected open spaces offer considerable resistance to ground-water flow and, therefore, have low permeability. The hydraulic characteristics of geologic materials vary between and within rock types. For example, in sedimentary deposits the permeability is a function of grain size and the range of grain sizes (the degree of sorting). Coarse, well-sorted gravel has much higher permeability than well-sorted sand. A well-sorted sand or gravel has a higher permeability than a deposit that is poorly sorted and has the open spaces between pebbles or sand grains filled with silt or clay. Clay-rich deposits generally have very low permeability. The permeability of lava flows also can vary markedly depending on the degree of fracturing. The highly fractured, rubbly zones at the tops and bottoms of lava flows and in interflow zones are often highly permeable, whereas the dense interior parts of lava flows can have very low permeability. Weathering and secondary mineralization, which often are a function of the age of the rock, can strongly influence permeability. Sedimentary deposits or lava flows in which the original open spaces have been filled with secondary minerals can have very low permeability.
Geologic properties that influence the movement of ground water within a flow system also can define the boundaries of the system. Geologic terranes consisting of predominantly low-permeability materials can form the boundaries of a regional flow system.
The upper Klamath Basin has been a region of volcanic activity for at least 35 million yr (years) (Sherrod and Smith, 2000), resulting in complex assemblages of volcanic vents and lava flows, pyroclastic deposits, and volcanically derived sedimentary deposits. Volcanic and tectonic processes have created many of the present-day landforms in the basin. Glaciation and stream processes have subsequently modified the landscape in many places.
The upper Klamath Basin lies within two major geologic provinces, the Cascade Range and the Basin and Range Province (Orr and others, 1992). The processes that have operated in these provinces have overlapped and interacted in much of the upper Klamath Basin. The Cascade Range is a north-south trending zone of compositionally diverse volcanic eruptive centers and their deposits extending from northern California to southern British Columbia. The Cascade Range is subdivided between an older, highly eroded Western Cascades, and a younger, mostly constructional High Cascades. Prominent among the eruptive centers in the High Cascades of the Klamath Basin are large composite and shield volcanoes such as Mount Mazama (Crater Lake), Mount McLoughlin, and Medicine Lake Volcano. The Cascade Range has been impinged on its eastern side by the adjacent structurally dominated Basin and Range Province. The Basin and Range Province is a region of crustal extension characterized by subparallel, fault-bounded, down-dropped basins separated by fault-block ranges. Individual basins and intervening ranges are typically 10–20 mi across. The Basin and Range Province encompasses much of the interior of the Western United States, extending from central Oregon southward through Nevada and western Utah, into the southern parts of California, Arizona, and New Mexico. Although the Basin and Range Province is primarily structural, faulting has been accompanied by widespread volcanism.
The oldest rocks in the upper Klamath Basin study area are part of the Western Cascades subprovince and consist primarily of lava flows, andesitic mudflows, tuffaceous sedimentary rocks, and vent deposits. The Western Cascade rocks range in age from 20 to 33 million years (my) and are as much as 20,000 ft thick (Hammond, 1983; Vance, 1984). Rocks of the Western Cascades overlie pre-Tertiary rocks of the Klamath Mountains Province, just west of the study area. Western Cascades rocks have very low permeability because the tuffaceous materials are mostly devitrified (changed to clays and other minerals), and lava flows are weathered and contain abundant secondary minerals. Because of the low permeability, ground water does not easily move through the Western Cascades rocks, and the unit acts as a barrier to regional ground-water flow. The Western Cascades constitute part of the western boundary of the regional ground-water flow system. Western Cascade rocks dip toward the east and underlie the High Cascade deposits, and define the lower boundary of the regional flow system throughout that part of the study area.
The High Cascade subprovince ranges in age from late Miocene (7 my) to Recent; however, most rocks are Pliocene (5 my) to Recent in age (Mertzman, 2000). Deposits within the High Cascade subprovince in the study area mostly form constructional features and consist of volcanic vents and lava flows with relatively minor interbedded volcaniclastic and sedimentary deposits. An area of numerous late Miocene to Pliocene cinder cones extends from southwest of Butte Valley to northwest of Mt. Mazama (Crater Lake). Quaternary volcanic deposits are associated with a few volcanic centers concentrated in two general areas in the upper Klamath Basin: from Lake of the Woods north to Crater Lake and from Mt. Shasta (south of the study area) east to Medicine Lake Volcano. The High Cascades rocks are relatively thin in southern Oregon and northern California, typically measured in hundreds of feet thick, rather than thousands (Stan Mertzman, Franklin and Marshall College, written commun., 2003). High Cascade rocks unconformably overlie Western Cascade rocks and are very permeable, relative to the older rocks.
Basin and Range Province deposits in the study area range in age from middle Miocene (13 my) to Recent. The oldest rocks are middle to late Miocene in age, ranging from 13 to 8 my. These rocks are exposed just south of the study area in the Pit River Basin and are equivalent to the upper Cedarville Series of Russell (1928). In the study area, those rocks probably underlie the Pliocene age lavas south of Clear Lake Reservoir. The older rocks in the Pit River Basin and bounding the eastern part of the study area are mostly silicic domes, flows, and pyroclastic deposits, which generally have low permeability (California Department of Water Resources, 1963) and typically are faulted and tilted.
Late Miocene to Pliocene volcanic rocks of the Basin and Range Province are the major water bearing rocks in the upper Klamath Basin study area. These units consist of volcanic vent deposits and flow rocks throughout the area east of Upper Klamath Lake and Lower Klamath Lake, and probably underlie most of the valley- and basin-fill deposits in the study area. Late Miocene to Pliocene rocks also form uplands along the eastern boundary of the study area, and form the plateau that extends from the Langell Valley south to the Pit River. The rocks are predominately basalt and basaltic andesite in composition, but silicic vents and lava flows occur locally, notably in the vicinity of Beatty, Oregon.
Tuff cones and tuff rings are the predominant volcanic vent form in the Sprague River subbasin between Chiloquin and Sprague River, Oregon. Tuff cones and rings form when rising magma comes in contact with water, resulting in explosive fragmentation of the volcanic material. The late Miocene to Pliocene rocks typically exhibit high to very high permeability. However, the permeability locally may be markedly reduced by secondary mineralization from hydrothermal alteration.
The volcanic rocks of the Basin and Range Province are interbedded with, and locally overlain by, late Miocene to Pliocene sedimentary rocks. The sedimentary rocks consist of tuffaceous sandstone, ashy diatomite, mudstone, siltstone, and some conglomerates. These units are exposed both in down-dropped basins and in up-thrown mountain blocks, indicating that the deposits in part represent an earlier generation of sediment-filled basins have been subsequently faulted and uplifted. These sedimentary deposits are typically poor water producers, and often serve as confining layers for underlying volcanic aquifers.
The youngest stratigraphic unit in the upper Klamath Basin consists of late Pliocene to Recent sedimentary deposits. Those deposits include alluvium along modern flood plains, basin-fill deposits within active grabens, landslide deposits, and glacial drift and outwash. Very thick accumulations of silt, sand, clay, and diatomite underlie the westernmost basins, such as the Upper Klamath Lake, Lower Klamath Lake, Butte Valley, and Tule Lake subbasins. For example, up to 1,740 ft of basin-fill sediment underlies the town of Tulelake, California. Sediment near the base of the deposit at Tulelake has been assigned an age of 3.3 my on the basis of radiometric ages of interbedded tephra, paleomagnetic data, and estimates of sedimentation rates (Adam and others, 1990). Gravity data suggest that the sediment-fill thickness may exceed 6,000 ft in the Lower Klamath Lake subbasin and may be in the range of 1,300 to 4,000 ft in the Upper Klamath Lake subbasin (Sammel and Peterson, 1976; Veen, 1981; Northwest Geophysical Associates Inc., 2002).
Hundreds of distinct and mappable geologic units have been identified by geologists in the upper Klamath Basin. Many of these geologic units have very similar hydrologic characteristics. For purposes of the description and analysis of regional ground-water flow, geologic units are typically combined into a smaller number of hydrogeologic units. Hydrogeologic units consist of groupings of geologic units that contain rock types of similar hydrologic characteristics and are distinct from other units. The geology of the upper Klamath Basin is herein generalized into eight hydrogeologic units (fig. 4 and table 1). Pre-Tertiary rocks are not exposed in outcrops or penetrated by wells in the study area and are not discussed.
Early to mid-Tertiary volcanics and sediments (Tovs), the oldest hydrogeologic unit in the study area, comprises Miocene and older lava and volcaniclastic rocks of the Western Cascade subprovince along the western margin of the study area, as well as older volcanic deposits beneath late Tertiary lavas along the eastern margin. The unit also includes older rocks exposed in the Pit River Basin southeast of the study area. The permeability of this unit is generally low due to weathering, hydrothermal alteration, and secondary mineralization. This unit is herein considered a boundary to the regional ground-water system of the upper Klamath Basin.
Late Tertiary volcaniclastic deposits (Tvpt) include palagonitized basaltic ash and lapilli deposits associated with eruptive centers. The hydrologic characteristics of this unit are not well known, but springs emerge from basal contact with unit Ts. This unit is most prominent in the Sprague River Valley.
Late Tertiary sedimentary rocks (Ts) consist predominately of fine-grained continental sedimentary deposits that include bedded diatomite, mudstone, siltstone, and sandstone. This unit has generally low permeability. These deposits occur throughout the central part of the upper Klamath Basin. They are exposed in uplands in interior parts of the basin and penetrated by wells in the river valleys. Lithologic logs of wells in the Sprague River Valley indicate that the thickness of these sedimentary deposits there locally exceeds 1,500 ft.
Late Tertiary volcanic rocks (Tv) consist predominately of basaltic and andesitic lava flows and vent deposits, but the unit includes local silicic domes and flows. This unit is locally affected by hydrothermal alteration and secondary mineralization. This is the most geographically extensive hydrogeologic unit, occurring throughout most of the upper Klamath Basin. The unit has moderate to high permeability and is by far the most widely developed aquifer unit in the study area.
Quaternary to late Tertiary sedimentary rocks (QTs) consist of medium- to coarse-grained unconsolidated to moderately indurated sedimentary deposits. The hydraulic characteristics of this unit are not well known, but lithologic descriptions on maps suggest that it is moderately permeable at some locations. This unit occurs locally in the western Wood River Valley, south of Klamath Falls, and in the uppermost Williamson River subbasin.
Quaternary volcanics (Qv) consist primarily of basaltic and andesitic lavas and vent deposits occurring in the Cascade Range and around Medicine Lake Volcano. These materials are generally highly permeable.
Quaternary volcaniclastic deposits (Qvp) consist primarily of pyroclastic flows and air-fall material (pumice ash and lapilli) deposited during the climactic eruption of Mt. Mazama that formed the caldera encompassing Crater Lake. This unit is most extensive in the Cascade Range around Crater Lake and in the upper Williamson River subbasin. As mapped (fig. 4), the unit also includes debris avalanche deposits in the Shasta River Valley outside of the study area. Minor Quaternary pyroclastic deposits occur on Medicine Lake Volcano and in Butte Valley. Air-fall deposits are highly permeable.
Quaternary sediments (Qs) include the alluvial deposits in principal stream valleys, glacial deposits in the Cascade Range, and basin-filling sediments in the major lake basins. The basin-filling deposits are generally fine grained and have low permeability. Coarse facies occur at some locations within the basin-filling deposits.
Geologic structures, principally faults and fault zones, can influence ground-water flow. Fault zones can act as either barriers to or conduits for ground-water flow, depending on the material in and between the individual fault planes. Faults most commonly affect ground-water flow by juxtaposing rocks of contrasting permeability or by affecting the patterns of deposition. Structural basins caused by normal faulting, called grabens, can act as depositional centers for large thicknesses of sediment or lava that may influence regional ground-water flow. Faults do not always influence ground-water flow; there are regions in the upper Klamath Basin where ground-water flow appears unaffected by the presence of faults.
The area of the upper Klamath Basin lying east of the Cascade Range is a composite graben that forms the westernmost structural trough of the Basin and Range physiographic province (Sherrod and Pickthorn, 1992). The predominant fault direction is north-northwest, as shown in figure 4. According to Sherrod and Pickthorn (1992), offset across the faults range from less than 300 ft in the central and eastern parts of the graben to about 6,000 ft on faults southwest of Klamath Falls.
Geologic materials possess certain hydraulic characteristics that control the movement and storage of ground water. This section describes the basic parameters used to characterize aquifer hydraulic properties and presents estimates or ranges of values of those terms for some of the major geologic units in the upper Klamath Basin. A more thorough discussion of the terms used to describe the hydraulic characteristics of aquifers and aquifer materials can be found in any basic ground-water hydrology text such as Freeze and Cherry (1979), Fetter (1980), or Heath (1983).
The term “permeability” was introduced previously as a measure of the ease with which fluid can move through a particular rock type or deposit. Permeability is an intrinsic property of the rock type, and is independent of the fluid properties. In ground-water studies, the term “hydraulic conductivity” is used more commonly than “permeability” in quantitative discussions. The hydraulic conductivity includes both the properties of the rock (the intrinsic permeability) and the properties of the water, such as viscosity and density. Hydraulic conductivity is generally defined as the volume of water per unit time that will pass through a unit area of an aquifer material in response to a unit hydraulic head gradient. Hydraulic conductivity has the units of volume per unit time (such as cubic feet per day) per unit area (such as square feet), which simplifies by division to length per unit time (such as feet per day). Hydraulic conductivity values for aquifer materials commonly span several orders of magnitude from less than 0.1 ft/d for fine sand and silt to over 1,000 ft/d for well-sorted sand and gravel.
When discussing aquifers instead of rock types, the hydraulic conductivity is multiplied by the aquifer thickness resulting in a parameter known as “transmissivity.” Transmissivity is defined as the volume of water per unit time that will flow through a unit width of an aquifer perpendicular to the flow direction in response to a unit hydraulic head gradient. Transmissivity has units of volume per unit time (such as cubic feet per day) per unit aquifer width (such as feet), which simplifies to length squared per unit time (such as feet squared per day [ft2/d]).
Storage characteristics of an aquifer are described by a parameter known as the “storage coefficient.” The storage coefficient is defined as the volume of water an aquifer releases from, or takes into, storage per unit area of aquifer per unit change in head. The volume of water has units of length cubed (such as cubic feet), the area has units of length squared (such as square feet), and the head change has units of length (such as feet). Thus, the storage coefficient is dimensionless. Storage coefficients typically span several orders of magnitude from 10-4 for aquifers with overlying confining units, to 0.1 for unconfined aquifers. Storage coefficients commonly fall between these two end members because aquifers often have varying degrees of confinement. Note that characterizing an aquifer as “confined” does not imply that it is not hydraulically connected to other aquifers or to surface water. The terms “confined” and “unconfined” describe the physics of the aquifer response to pumping at a particular location.
The hydraulic characteristics of subsurface materials are typically determined by conducting aquifer tests. An aquifer test consists of pumping a well at a constant rate and measuring the change in water level (the drawdown) with time in the pumping well and nearby non-pumping wells. The data collected allow generation of a curve showing the drawdown as a function of time. Similar data are collected after the pumping is stopped, allowing generation of a curve showing the water-level recovery as a function of time. Analysis of the drawdown and recovery curves in the pumped well and observation wells provides estimates of the transmissivity and storage coefficient of the aquifer. Aquifer characteristics also can be estimated from certain well-yield tests called “specific-capacity tests,” sometimes conducted by drillers. Data from specific capacity tests that include a pumping rate, test duration, drawdown at the end of the test, and the well diameter can be used to estimate aquifer transmissivity.
The results of 32 aquifer tests conducted in the upper Klamath Basin are summarized in table 2. The tests were conducted by the OWRD, the California Department of Water Resources (CDWR), private consultants, and the USGS. Pumping periods for the tests ranged from 12 hr (hours) to 169 days, with most lasting 24 to 72 hr. All tests were conducted on wells with large yields ranging from about 1,000 to 10,000 gal/min. Most tests are of wells that produce from Tertiary volcanic deposits (unit Tv on fig. 4) because it is the most productive and widespread water bearing unit. A smaller number of tests were of wells producing from Tertiary sedimentary deposits (unit Ts on fig. 4) or a mixture of Ts and Tv. It should be noted that the Tertiary sediments are very fine grained over most of the basin, and that wells producing large yields from that unit occur only in specific locations.
Data and details of the analyses for most of the tests are available from sources listed in table 2. Reanalysis of the aquifer tests listed in table 2 was beyond the scope of this study. For the most part, the results presented are directly from the source documents, except that values have been rounded to two significant figures. In some cases, as noted in table 2, results from certain observation wells or certain analyses that were considered problematic were not included. For example, anomalous results from observation wells that were open to different water-bearing zones or constructed differently from the pumped well were excluded. Results from pumped wells were excluded where well loss (excessive drawdown due to well inefficiency) appeared to affect the results.
Most aquifer tests show evidence of boundaries, complicated aquifer geometry, or possible double-porosity conditions where flow occurs in fractures and in the blocks between fractures. Many tests in Butte Valley and the Tule Lake, Lower Klamath Lake, Sprague River, and upper Lost River subbasins showed inflections in drawdown curves, suggesting the presence of no-flow boundaries. These no-flow boundaries were in some cases associated with faults. Such boundaries indicate that the Tertiary volcanic aquifer system is, at least locally, somewhat compartmentalized, with some resistance to flow between individual subregions. Some tests showed evidence of recharge boundaries. Recharge boundaries usually indicate that the cone of depression has expanded to an extent where it has intersected a source of recharge, for example a stream or canal. Given the stratigraphy of the areas tested, the pumping more likely was inducing flow from the overlying low-permeability sediments. Tests that showed evidence of recharge boundaries or leaking confining layers occurred in the Lower Klamath Lake and Lost River subbasins. Inflections in drawdown curves can also be caused by double porosity conditions (Moench, 1984; Weeks, 2005).
Aquifer tests show that the transmissivity of the Tertiary volcanics (predominantly basaltic lavas) varies widely, from 2,700 to 610,000 ft2/d, with most (the middle 50 percent) ranging from 24,000 to 270,000 ft2/d. The average transmissivity is about 170,000 ft2/d and the median is about 90,000 ft2/d. Reported storage coefficients from aquifer tests in the Tertiary volcanics range from 0.00001 to 0.15. The 0.15 figure is anomalous and likely due to a partially penetrating observation well and leakage from the confining layer. The middle 50 percent of the calculated storage coefficients in the Tertiary volcanics range from 0.00025 to 0.001. The average value is 0.0012 and the median is about 0.0005.
Although the number of aquifer tests in Tertiary sediments (or mixtures of the sediments and Tertiary lavas) is small (n=6, not including the geothermal aquifer test), they provided information on the hydraulic characteristics of the coarse-grained facies of unit Ts. Transmissivity values range from 13,000 to 350,000 ft2/d, with most in the 25,000 to 75,000 ft2/d range. The average value is about 100,000 ft2/d and the median is 54,000 ft2/d. Storage coefficients range from 0.0005 to 0.015 with most ranging from about 0.0002 to 0.003. Note that most Tertiary sedimentary rock in the basin consists of fine-grained lake deposits and has much lower transmissivity than determined from the tests discussed here.
In the early 1980s, the USGS conducted an aquifer test of the geothermal aquifer in Klamath Falls in collaboration with the Lawrence Berkeley Laboratory and the City of Klamath Falls (Benson and others, 1984a, b). The test consisted of four phases: a 1-week pre-test phase during which background water levels were monitored; a 21-day pumping phase during which a geothermal well (38S/09E-28DCB, KLAM 12050; table 2) was pumped at about 720 gal/min and the water discharged to an irrigation canal; a 30 day injection phase during which pumping continued (at about 660–695 gal/min) and the water injected into a second well (38S/09E-28DDD, KLAM 11940); and a 1-week recovery phase. Benson and others (1984a) analyzed the data from the test and calculated a permeability-thickness value (analagous to a transmissivity) of about 1.4 X 106 millidarcy-feet. This converts to a transmissivity of about 3,800 ft2/d. Analysis of the test indicated a storage coefficient of about 0.002.
Results of the geothermal aquifer test are generally consistent with the other aquifer tests in table 2. The transmissivity value is at the lower end of the range of other tests, but this is not unexpected as the aquifer system pumped consists of interlayered lava and fine-grained sedimentary rock (unit Ts). A notable finding of this test is the apparent lack of boundaries encountered in an area crossed by several major basin-bounding faults. This is not, however, inconsistent with other hydrologic data that suggest ground water moves freely across similar faults at many locations.
Another source of information on subsurface hydraulic characteristics are the well-yield tests conducted by drillers and reported on the well logs submitted on completion of all new wells. Well-yield tests typically consist of a single drawdown measurement taken after a well has been pumped at a specified rate for a specified length of time, typically 1 hr. Well-yield tests allow determination of a well’s specific capacity, which can be used to estimate transmissivity as described previously. Specific capacity is only a semiquantitative measure of well performance in that it can vary with pumping rate. Specific-capacity values can be used to calculate only rough estimates of the aquifer transmissivity and cannot be used to quantitatively derive aquifer storage characteristics. Although transmissivity values calculated from specific capacity tests are only approximate, they can be used to evaluate the relative differences in hydraulic characteristics between different geographic areas and different hydrogeologic units if data are available from a sufficient number of wells.
Specific-capacity data were analyzed from wells that were field inventoried for this study. Of the over 1,000 wells inventoried, only about 288 had sufficient information for analysis on their State water well reports. Transmissivity values were estimated from specific-capacity data using the Theis nonequilibrium equation (Theis, 1935). The wells analyzed were sorted by hydrogeologic units for comparison. Most wells analyzed produced from one of three units: Quaternary sedimentary deposits (Qs) (n = 41), Tertiary sedimentary rocks (Ts) (n = 48), and late Tertiary volcanic deposits (Tv) (n = 173). Other units had too few tests for statistically meaningful comparisons. The cumulative frequencies of transmissivity estimates for the three major units are shown in figure 5. Wells producing from Quaternary sedimentary deposits and Tertiary sedimentary deposits have similar transmissivity distributions, with the former having slightly larger values. The median transmissivity for both units is about 200 ft2/d. The frequency distribution of transmissivities for the late Tertiary volcanic deposits is distinct from the other units, with values generally larger by more than an order of magnitude (fig. 5). The median transmissivity of Tertiary volcanic deposits is about 5,800 ft2/d.
The median transmissivity for late Tertiary volcanic deposits determined from specific-capacity tests (6,300 ft2/d) is lower than that calculated from aquifer tests (about 90,000 ft2/d). This is not unexpected for the following reasons: First, transmissivity values determined from single-well tests can be biased downward by excess drawdown in the pumped well due to well inefficiency (see Driscoll, 1986, p. 244). Aquifer tests with observation wells are not affected by this phenomenon. Second, the large number of specific-capacity tests (173) represent a more or less random sampling of wells (and varying characteristics) in the unit. Aquifer tests, in contrast, are not random but tend to be conducted most commonly on high yielding wells for specific purposes. Regardless, transmissivity values calculated from both aquifer tests and specific-capacity tests are useful for understanding the hydraulic characteristics of hydrogeologic units and the differences between units.
Back to Table of Contents