WRIR 01-4195:
Ground-Water Discharge Determined from Estimates of Evapotranspiration,
Death Valley Regional Flow System, Nevada and California
Previous section: General Description and Setting | Next section: Summary | Table of Contents | Conversion Factors and Acronyms | Report Home Page
Historically, estimates of evapotranspiration in the DVRFS were computed as part of regional assessments of the ground-water resource (Malmberg and Eakin, 1962; Walker and Eakin, 1963; Pistrang and Kunkel, 1964; Malmberg, 1967; Rush, 1968). These regional assessments estimated annual ET losses as the product of the acreage of phreatophytes within a discharge area and an annual ET rate representative of the vegetation and soil conditions of the discharge area. Acreages were estimated by delineating vegetation from aerial photographs and field mapping. Annual ET rates typically were estimated from values reported for similar plant assemblages found throughout the western United States (Lee, 1912; White, 1932; Young and Blaney, 1942; Gatewood and others, 1950; Robinson, 1958). Most of these rates were determined from tank experiments as described by Gatewood and others (1950, p. 105-111). Although the methodology is technically sound, its overall accuracy depends on the accuracy of the individual components. Inexact estimates of the acreage, the annual ET rate, or both could introduce substantial error into the calculation. Recent studies of ET rates for vegetation and soil conditions in and near the DVRFS (Johnson, 1993; Czarnecki, 1997; Laczniak and others, 1999; Nichols, 2001) indicate rates somewhat different than those presented in earlier reconnaissance studies. These more recent estimates of ET rates were determined primarily from energy-balance techniques (Brutsaert, 1982) that require rigorous and long-term measurements of micrometeorological data. Noted differences between early and more recent measurements of ET rates are expected considering differences in the local climatic, vegetation, and soil conditions between sites and differences in the methods used to measure ET.
Most previous attempts to quantify ground-water discharge from the DVRFS have been based on measurements of springflow or on estimates of ET from the major discharge areas (Malmberg and Eakin, 1962; Walker and Eakin, 1963; Pistrang and Kunkel, 1964; Malmberg, 1967; Rush, 1968, Dudley and Larson, 1976). In discharge areas where seeps are dominant or where springs are difficult to measure, estimates of ground-water discharge account only for measurable outflow. Alternatively, ground-water discharge has been estimated by efforts to quantify local ET. Assuming that all spring and seep flow is evaporated or transpired by the local vegetation, ET-based estimates also include water upwardly diffused into the shallow flow system from the underlying regional flow system.
The need for accurate estimates of ground-water discharge and the differences in ET rates estimated by past and more recent studies prompted studies to re-evaluate and more rigorously and consistently quantify mean annual ET and ground-water discharge throughout the DVRFS. The approach used in this study is similar to that used in the early reconnaissance studies in that ET is calculated as the product of an area and a rate. Although the approach is similar, the specific techniques used to delineate ET areas and to determine an appropriate ET rate differ in that they take advantage of modern technologies and are applied consistently over the DVRFS. Because these techniques are expected to reduce errors, the revised ET estimates should be more reliable. Early reconnaissance studies delineated ET areas based on generalized vegetation and soil mapping -- this study utilizes satellite imagery in combination with remote sensing and a geographic information system. The reconnaissance studies estimated ET rates based on measurements made for similar phreatophytes found at locations outside the study area -- this study relies primarily on recent estimates determined from micrometeorological measurements made in areas within or adjacent to the study area (Johnson, 1993; Czarnecki, 1997; Laczniak and others, 1999, and Nichols, 2001; and Reiner and others, 2001). Another difference between the two approaches is in the calculation of ground-water discharge. The reconnaissance studies assume that ground-water discharge is equivalent to ET; this study, however, estimates discharge by reducing the ET rate to account for any local precipitation falling on the area. This reduction to the ET rate generally is small considering the aridity of the area.
ET and ground-water discharge estimates computed in this study account only for ground water lost to the atmosphere. Estimates do not account for ground water pumped from the flow system or for ground water exiting the discharge area as underflow. Presently, ground water consumed to support agricultural, domestic, and operational needs is minimal and probably does not exceed more than a few hundred acre-feet per year in any major discharge area other than in Pahrump and Lida Valleys. In Pahrump Valley, about 30,000 acre-ft of ground water was pumped for municipal and domestic supply and irrigation in 1997 (Division of Water Resources, State of Nevada, written commun., 1999). In Lida Valley ground water used for irrigation is less than a few hundred acre-feet per year but accounts for a significant portion of the ground water discharged from that area. Because the focus of this study is on natural ground-water discharge, and the difficulty associated with separating transpiration by natural vegetation from that of irrigated crops, ET and ground-water discharge estimates for Pahrump and Lida Valleys are not included in this report.
Annual ET is estimated for Chicago Valley, the Franklin Well area, Franklin Lake, Sarcobatus Flat, the Shoshone area, Stewart Valley, and the Tecopa/California Valley area (fig. 1). Also included in this report are estimates for Ash Meadows and Oasis Valley, which were studied primarily as part of previous efforts (Laczniak and other, 1999; and Reiner and others, 2002, respectively). Although Death Valley is an area of substantial natural ground-water discharge, it is not included because evapotranspiration rates for salt-encrusted playas are unavailable. Evapotranspiration rates in Death Valley are being studied by the U.S. Geological Survey in cooperation with the National Park Service as part of an ongoing study. Ground water removed by processes other than ET from the major discharge areas included in this report is assumed to account for less than a few percent of the total discharge.
Past studies have shown that ET rates throughout the Great Basin region vary with vegetation and soil conditions. In general, the more dense and healthy the vegetation and the wetter the soil, the greater the rate of ET (Ustin, 1992; Nichols, 2001). Vegetation, water, and soil covers reflect incoming solar radiation differently. These differences have been used to determine type, density, and health of vegetation, and type and moisture content of soil from aerial and satellite imagery (Anderson and others, 1976, p. 2; American Society of Photogrammetry, 1983, p. 23-25; Goetz and others, 1983, p. 576-581). This process is referred to as a land-cover classification. The procedure used in this study takes advantage of (1) the relation between ET, and vegetation and soil conditions, and (2) differences in the spectral reflectance of the different vegetation and soil covers to map distinct ET units within the major discharge areas of the DVRFS, and hereafter is referred to as an ET-unit classification.
Thematic Mapper (TM) imagery was the source of the spectral data used to classify ET units. Earth-orbiting satellites equipped with sensors to detect solar-reflected and earth-emitted radiation acquire TM imagery. TM satellites collect spectral information across seven wavelength bands referred to as TM channels (fig. 4). Each channel spans a discrete part of the electromagnetic spectrum. Six of the channels (1, 2, 3, 4, 5, and 7) measure reflected radiation in the visible, near infrared, and short-wave infrared regions. These six channels were used in this study to define the spectral character or response of an area imaged by the satellite. A seventh band, TM channel 6, measures thermal energy radiated from the Earth and was not used in this study. Spectral reflectance, as acquired by TM sensors, represents an average value over an area measuring about 100 ft by 100 ft. Each square area is referred to as a pixel (picture element) and defines the spatial resolution of the imagery. Differences in the spectral response of land covers having different vegetation, soil, and moisture conditions are shown in figure 4A.
Figure 4B shows these same spectral responses as would be inferred from TM imagery. The TM imagery used to classify all discharge areas, except Sarcobatus Flat, was acquired June 13, 1992. The decision to use June 1992 imagery was based on (1) June typically being a period of high vegetation vigor, and (2) 1992 having slightly above normal in terms of precipitation. Two TM scenes (entity-id number LT5040034009216510 and LT5040035009216510) were required for complete coverage of the area of interest. A pseudo-color infrared composite of the combined TM imagery is shown in figure 5. Excessive cloud cover in the June 13, 1992, imagery precluded accurate classification of ET units in the Sarcobatus Flat area. Instead, the Sarcobatus Flat area was classified from imagery acquired June 21, 1989 (LT5040034008917210). One major difference between the procedure used here and that used by Laczniak and others (1999) for the Ash Meadows area was in the number of dates used for the classification process. Here only June imagery was used to classify ET units, whereas, Laczniak and others used both imagery from June and September 1999. The June imagery represented conditions of near maximum plant vigor and high moisture, whereas the September imagery represented conditions of high plant stress (dormancy) and low moisture. Based on the Ash Meadows results, a single-date classification was assumed adequate for the discharge areas being evaluated.
The procedure used to classify ET units applied an unsupervised approach (Avery and Berlin, 1992, p. 452) to identify the unique spectral responses present within the TM imagery. Uniqueness of the spectral response was based on statistical differences between reflectance values in TM channels 1, 2, 3, 4, 5, and 7 of the imaged pixels falling within the boundaries of the major discharge areas (fig. 1). Boundaries initially were located using a raster image representing a modified soil adjusted vegetation index (Qi and others, 1994) developed from June 13, 1992, TM imagery. Boundaries were refined continually on the basis of information acquired during numerous field visits. The use of boundaries reduced the area within the TM image that needed to be classified to that of the major discharge areas. Final boundary locations are available in digital format from the USGS node of NSDI at http://water.usgs.gov/lookup/getspatial?darea.
Each pixel was assigned a unique spectral response using the maximum likelihood classification technique (Lillesand and Kiefer, 1987, p. 685-689). This technique compares reflectance values of each pixel with the statistics defining each unique spectral response, calculates the statistical probability of a pixel being associated with each unique spectral response, and assigns each pixel to the unique response having the highest probability.
The classification process then reduces the number of unique spectral responses by grouping them on the basis of similarities in their reflectance statistics. Each group typically represents a unique land cover. Groupings were further lumped into clusters on the basis of similarities in vegetation and soil conditions. The consolidation process was an interactive procedure that involved comparing similarities in the spectral response with information gathered in the field. The process continued until a manageable number of clusters remained whereby each represented a distinct ET unit. This process ultimately resulted in 10 clusters, each representing a unique area of significant ground-water ET typified by a land cover dominated by open water, phreatophytes, or moist bare soil (table 1; fig. 6). An eleventh cluster included all spectral responses representing areas of insignificant ground-water ET typified by land covers dominated by sparse upland desert vegetation and xeric landscape. Each cluster was given a number from 0 to 10 (table 1; fig. 6).
Most spectral responses included within a cluster exhibit a characteristic shape (fig. 6). The two exceptions are clusters 4 (dense meadow and forested vegetation) and 7 (moist bare soil). Each of these clusters is defined by spectral responses exhibiting multiple characteristic shapes. The multiple patterns within each of these clusters are explained by the inclusion of more than one vegetation or soil type within the cluster. For example, cluster 4 includes dense meadow and forested vegetation, each of which exhibits a different spectral response. The two spectral responses are combined into one cluster because ET rates between the vegetation types are assumed to be similar.
The final step in the classification procedure is to assign an ET unit to each pixel. Pixels outside the boundary of a major discharge area were assigned a value of zero (an area of insignificant ground-water ET). Pixels within discharge areas were assigned a value of 1 through 10 to represent their associated cluster. This process created a raster image whereby each pixel within the image was associated with a single ET unit. The image was resampled at a finer resolution (60 ft by 60 ft) to allow for direct comparison with results presented for Ash Meadows (Laczniak and others, 1999) and Oasis Valley (Reiner and others, 2002). The resampled image was then smoothed using a filter, which replaced spuriously classified pixels (areas defined by less than three adjacent pixels) in the image and filled single-pixel gaps by assigning the pixel to the ET unit most representative of its neighbors.
The spatial distribution of ET units as classified by the above procedure was mapped for each major discharge area (figs. 7-15). The classification shown for Ash Meadows (fig. 7) is that given by Laczniak and others (1999) with the addition of one ET unit to represent open playa. Open playa was added to maintain consistency with other mapped discharge areas. Raster data sets representing the final classification of each discharge area are available from the USGS node of NSDI at http://water.usgs.gov/lookup/getspatial?etunit. The acreage of each ET unit and the total ET-unit acreage are given for each discharge area in table 2.
The discharge area having the largest ET unit acreage is Sarcobatus Flat at 34,250 acres (table 2; fig. 12). Of this total, the majority is classified as sparse to moderately dense shrubs at 19,372 acres or open playa at 10,817 acres (tables 1 and 2). The dominant phreatophyte found in the shrubland area of Sarcobatus Flat is greasewood (Sarcobatus vermiculatus). The discharge area having the next largest acreage is Ash Meadows at 12,467 acres (table 2; fig. 7). This area is about 2,100 acres greater than that given by Laczniak and others (1999, table 10). This difference results from a slightly different interpretation of the Ash Meadows boundary and the inclusion of 2,241 acres of open playa as an ET unit. The largest ET unit classified within Ash Meadows is sparse grassland vegetation at 7,059 acres (tables 1 and 2). This grassland vegetation is dominated by expansive meadows of saltgrass (Distichlis spicata var. stricata). The area for sparse grassland vegetation and moist bare soil is about 150 acres less than that reported in Laczniak and others (1999, table 10). This area was not overlooked but is instead included in the adjacent Franklin Lake discharge area (figs. 1 and 10). Together these two disparities account for the difference in total acreage between the two reports. The discharge area having the smallest total acreage is the Franklin Well area at 297 acres (table 2; fig. 9). This discharge area is a narrow crescent shape that spans only about a 5-mi section of the Amargosa River along the Nevada-California border (fig. 1).
The primary sources of ET rates were from recent studies of Ash Meadows (Laczniak and others, 1999, table 7) and Oasis Valley (Reiner and others, 2002). As part of these two studies, 15 sites were instrumented to collect micrometeorological data (table 3; figs. 7 and 11). These data were used to quantify ET rates for most of the ET units found within the discharge areas of interest. Daily ET rates were computed using the Bowen ratio solution of the energy-budget equation (Bowen, 1926). ET sites were instrumented from 1 to 3 years, and annual ET rates were determined from the data collected.
ET rates for each ET unit are presented as ranges in table 4. The range given for each ET unit is inclusive of all ET rates computed for Ash Meadows (Laczniak and others, 1999) and Oasis Valley (Reiner and others, 2002), and of rates estimated in other selected studies of ET throughout the general area. Data sources and the relative significance of each source used to construct the range are listed in table 4. The range includes ET rates computed for different sites during different years. The highest annual rate is near 9 ft for open water and submerged aquatic vegetation. The lowest annual rate is less than 1 ft for open playa. The difference between the minimum and maximum values varies among ET units, ranging from 0.4 ft/yr for open water and submerged aquatic vegetation to 1.8 ft/yr for sparse to moderately dense shrubs (table 4). Larger differences are associated with units having greater spatial variation in vegetation density.
The estimated ET rate for each ET unit within a discharge area is listed in table 5. Estimates represent the mean annual ET rate. Excluding open playa, ET rates given for Ash Meadows are those estimated by Laczniak and others (1999). Rates given for ET units in other discharge areas reflect differences between the different discharge areas in vegetation density. The density adjustment was made based on Ash Meadows values for ET units within Ash Meadows, and for other discharge areas on the mean value of the range.
Density differences between ET units were determined from a modified soil adjusted vegetation index (MSAVI; Qi and others, 1994). The MSAVI uses TM channels 3 and 4 to compute a vegetation index for each pixel in the imagery. The MSAVI increases the dynamic range of the vegetation signal by minimizing background influences from the soil. Index values range from 0 to 1, where 1 represents the strongest vegetation signal. A MSAVI was computed using the June 13, 1992 (fig. 16), and June 21, 1989, TM imagery. Raster data sets representing these indices are available from the USGS node of NSDI at http://water.usgs.gov/lookup/getspatial?msavi89 and http://water.usgs.gov/lookup/getspatial?msavi92. The average MSAVI computed for each ET unit is listed in table 6. Averages are based on an index computed from the imagery used to classify the discharge area, and range from 0.045 for submerged aquatic vegetation to nearly 0.395 for wetland vegetation. In general, lower values are associated with units having the sparsest vegetation and higher values with units having the densest vegetation. ET units having a greater average MSAVI were assigned higher ET rates, and those having a lesser average were assigned lower rates.
Estimates of the mean annual volume of ET from each ET unit and discharge area are listed in table 5. Volume estimates for each ET unit are computed as the product of a unit's acreage and rate, and the total for each discharge area is computed by summing the individual ET-unit estimates. Mean annual ET ranges from about 450 to 30,000 acre-ft. Discharge areas having the largest mean annual ET are Ash Meadows and Sarcobatus Flat at 22,000 and 30,000 acre-ft, respectively, and those having the smallest are the Franklin Well area and Chicago Valley at 450 and 650 acre-ft, respectively.
Sources contributing water to ET include direct precipitation, surface-water inflow from the surrounding drainage area, and regional ground-water inflow. Regional ground-water inflow includes recycled spring and seep flow and any diffuse upward flow from the underlying regional flow system. Assuming that all ground water discharged from the regional flow system is locally evaporated or transpired, ground-water discharge can be estimated from ET knowing the precipitation and surface-water inflow components. In this study ground-water discharge is computed assuming that (1) the surface-water contribution is negligible, and (2) the precipitation component of ET is approximately equal to precipitation falling on the area -- assumptions considered reasonable for these arid valleys. Spring and seep flow are not directly accounted for in ET-based estimates, but are indirectly included by the assumption that all local surface discharge (spring and seep flow) is evaporated or recycled back into the shallow ground-water flow system, where eventually, it is evaporated or transpired from within the area.
The average annual precipitation to each discharge area was estimated from precipitation measurements made at ET sites in Ash Meadows, Oasis Valley, and Death Valley (table 7); long-term continuous data collected at nearby National Weather Service (NWS) climate stations (tables 8 and 9); and values interpreted from different published maps of precipitation (Hardman, 1965; Houghton and others, 1975, fig. 40; Winograd and Thordarson, 1975, fig. 3) and a map generated by parameter-elevation regressions on independent slopes model (PRISM; Daly and others, 1994). Estimates range from 3.5 in. in the Shoshone and Tecopa/California Valley discharge areas to 6 in. in the Sarcobatus Flat and Oasis Valley discharge areas.
An estimate of mean annual ground-water discharge from each discharge area was computed by summing individual estimates of mean annual ground-water discharge computed for each ET unit within the discharge area (table 10). Mean annual ground-water discharge from each ET unit was computed as the product of an adjusted mean annual ET rate and the acreage of the ET unit. Adjustments were made to remove any water from the ET estimate that was contributed by precipitation. The remainder of the water consumed by ET is assumed to have originated from ground water. Mean annual ET rates (table 10) were adjusted for precipitation by subtracting an estimate of annual precipitation. Adjusted rates are listed in table 10 for each discharge area.
The estimate of mean annual ground-water discharge for each discharge area, determined from adjusted ET rates, is listed in table 10. Volumes range from 350 acre-ft for the Franklin Well area to 18,000 acre-ft for Ash Meadows. Differences between mean annual ET (table 5) and mean annual ground-water discharge (table 10) range from 100 acre-ft in the Franklin Well area to 17,000 acre-ft in Sarcobatus Flat. Differences on a percentage basis are largest in Sarcobatus Flat, Stewart Valley, and Franklin Lake at 57, 48, and 44 percent, respectively. The smallest difference is 14 percent in the Tecopa/California Valley area. The larger percent differences generally are associated with discharge areas dominated in part by open playa.
Estimates of mean annual ground-water discharge listed in table 10 differ some from previous estimates reported in the literature (Malmberg and Eakin, 1962, p. 16-17, 25; Walker and Eakin, 1963, p. 24; Malmberg, 1967, p. 29; Blankennagel and Weir, 1973, p. 21; Winograd and Thordarson, 1975, p. 84; Czarnecki and Waddell, 1984; Czarnecki, 1997, p. 58; D'Agnese and others, 1997, p. 46). For a particular discharge area, previous estimates often show a wide range between low and high values and may vary by a factor of two or more. This wide range makes it difficult to determine differences between estimates presented here and previous estimates. The determination is further complicated because previous estimates often are reported for large areas that include multiple discharge areas and the actual extent of individual discharge areas often differs from that defined in this report. In general, ground-water discharge estimates reported here are greater than those reported in the literature for the northern discharge areas (Sarcobatus Flat and Oasis Valley) and less than those reported in the literature for the southern discharge areas (Chicago Valley, Franklin Lake, Shoshone, and Tecopa/California Valley). These discrepancies relate to differences in estimates of ET acreage, ET rates, or both.
The accuracy of the ground-water discharge estimates is limited primarily by the assumptions inherent in the classification procedure used to define ET units and in the energy-budget methods (primarily Bowen ratio) used to compute ET rates (Laczniak and others, 1999, p. 21-22). Other factors potentially affecting the accuracy of the ground-water discharge estimates include (1) the assumption that all springflow ultimately is evaporated or transpired from within the bounds of defined discharge areas; (2) the assumption that no external surface inflow contributes water to the local ET process; (3) the short-term nature of the data used to compute mean annual ET rates; (4) the limited number of sites used to estimate mean annual ET rates; (5) uncertainty in estimates of ET rates based on computed relative density differences; and (6) the uncertainty in the adjustment applied to remove precipitation from ET estimates. Uncertainty associated with ground-water discharge estimates is determined using a Monte Carlo analysis. Results of the analysis are discussed in the appendix.
With one exception, ET units and their associated acreage were determined from TM imagery acquired on a single date (June 13, 1992). Cloud cover over Sarcobatus Flat on that date necessitated the use of TM imagery acquired on June 21, 1989, for that area. June represents a period of the year when vegetation is relatively robust, and 1992 was a year of slightly above normal precipitation. Vegetation on this date is assumed to be healthy and vigorous, and thus should be easily classified through spectral methods. Although this date is assumed to be a reasonable representation of average vegetation and soil conditions throughout the DVRFS, conditions are known to change from one year to the next (Laczniak and others, 1999, p. 31-33). Most of these changes are a consequence of changes in precipitation, which affects local vegetation, soil-moisture conditions, and the depth to the water table -- all of which affect ET rates. Thus, a classification technique based on multiple years of TM imagery or some other type of spectral data would provide more confidence in the classification of ET units and acreage calculations in terms of long-term averages.
Estimates of the ET rate for each ET unit were determined using rates computed from micrometeorological data collected at 15 sites instrumented throughout Ash Meadows and Oasis Valley, and from ET rates given in other selected studies of the general area. Together these rates established a range of values that generally define the ET rate for each ET unit. In some units, the range includes rates computed at multiple instrument sites (up to five), while in others, it includes only values reported in the literature. Some sites were instrumented for periods of up to 3 years, while others only for a single year. No sites were instrumented in the discharge areas of Chicago Valley, Franklin Lake, Franklin Well area, Sarcobatus Flat, Shoshone area, Stewart Valley, and Tecopa/California Valley area. ET rates determined over longer time periods and from additional sites in other discharge areas would help refine, improve, and provide more confidence in estimates of mean annual ground-water discharge. This is especially true of more sensitive units, such as open playa, where estimates of annual ET rates are based on limited local data and on values reported in the literature for vegetation and soil conditions outside the study area.
The ET rate assigned to each ET unit within a discharge area was selected from a range of values based on relative differences in vegetation density. The range is intended to be a general indicator of likely ET rates expected for a particular ET unit and its width defines the variability in the vegetation and moisture conditions within a unit. For example, a larger range suggests greater changes in vegetation density. Relative density differences between ET units from different discharge areas were determined from differences in the average MSAVI value computed from TM imagery. Although this technique is likely to provide a reasonable estimate of the ET rate, more accuracy could be achieved either by classifying more ET units or by improving the relation between density and ET rate. Either approach requires the installation of additional ET sites to measure local ET rates.
ET estimates were adjusted to remove contributions of local precipitation. Some uncertainty is inherent in the precipitation adjustment. This uncertainty is attributed to errors in estimating the average annual precipitation and to the uncertainty in the actual amount of local precipitation included in estimated ET rates. No adjustment was made to remove any surface-water inflow contribution from the estimate of the ET rate. The aridity of the area is likely to produce minimal surface-water inflows, but in discharge areas where precipitation is relatively high and the surrounding drainage area is dominated by highlands, such as Oasis Valley and Sarcobatus Flat, surface-water inflow may be a more substantial component of the estimated ET rate. The decision not to adjust ET rates for local surface-water inflow was based on the scarcity of available data and may be partly responsible for the larger ground-water discharge estimates given for the northern discharge areas than reported in the literature. Additional data defining the amount of water contributed by the many sources to local ET would greatly improve estimates of ground-water discharge.
Unclassified areas are assumed to be zones of no substantial ground-water discharge. This assumption, although strongly supported by the lack of vegetation, dryness of soil, and greater depths to the water table (generally exceeding 50 ft), could result in some error when estimating ground-water discharge. Even over vast areas, volumetric losses of ground water are likely to be negligible considering the very low ET rates (less than 1.0x10-4 ft/yr) associated with these areas (Andraski, 1997, table 2).
Ground-water discharge estimates as reported include only water lost through evaporation and transpiration, and do not include any water that may be leaving discharge areas through subsurface flow. Until additional data become available from which to define water-level distributions and spatial variations in the hydraulic properties controlling ground-water movement throughout the shallow local and deeper regional flow systems, reliable estimates of subsurface outflow are not possible. Considering the high potential for subsurface outflow from many of the major discharge areas in the DVRFS, the estimate of ground-water discharge presented should be considered a minimum value for the total amount of ground water exiting a discharge area.
Previous section: General Description and Setting |
Next section: Summary
Table of Contents | Conversion
Factors and Acronyms | Report Home Page