Evidence for binary piston-flow mixing
Field parameters (Table S1) and major-ion chemistry from both current (Table S2) and previously published studies (Metcalf 1995; Yelkin 1996; Schaefer et al. 2006; Hershey et al. 2010) suggest that water discharging from Littlefield Springs is a mixture of seepage from the Virgin River and groundwater from the regional Great Basin carbonate aquifer. The trilinear diagram (Figure 3) shows two potential end-member waters in the Virgin River Gorge subarea: (1) Virgin River samples collected at VR-8 and VR-11 upstream of Littlefield Springs, and (2) a synthetic composite representing the carbonate aquifer. This composite was based on samples from the McCullough Well (MCW) completed in the Permian Kaibab Formation about 30 km east of the study area (Wilkowske and Heilweil 1998), Blue Point Spring (BPS) and Rogers Spring (RGS) discharging from gypsiferous carbonates near Lake Mead about 60 km southwest of the study area (Hershey et al. 2010), and Lost Spring (LST) discharging from the Permian Kaibab Formation in the Virgin Mountains (this report; Figure 1). LST is considered a carbonate-aquifer end member for major and trace-ion geochemistry, but not for age or recharge conditions since it is located south of the springs rather than receiving water from the primary recharge area in the northern part of the study area. Hershey et al. (2010) interpreted BPS and RGS as located at the terminal end of the carbonate aquifer based on chemical evolution as groundwater moves through the system. The similarity in major- and trace-ion chemistry between these springs and Littlefield Springs implies interconnection and discharge from a shared regional carbonate aquifer system (Yelkin 1996). Littlefield Springs (LFS-1 through LFS-13) and the Virgin River at Desert Springs (VR-DS) lie at an intermediate location between these two end members, indicating a binary mixture. Similarly, Littlefield Springs water temperatures (18 to 27 °C; Table S1) lie between the reported 30 °C temperature for BPS and RGS (Hershey et al. 2010) and the 3-year mean Virgin River temperature of 17 °C measured at VR-8 (Martin Schijf, Utah Division of Wildlife Resources, personal communication, 2011). Unlike the other Virgin River samples, VR-DS was sampled when the Virgin River was dry above Littlefield Springs, resulting in a weighted average of all the spring discharge above this location. Major-ion chemistry indicates that water from the Cedar Pocket Well (CPW), located adjacent to the Virgin River and screened from 80 to 110 m, is similar to VR-15 and likely is hydraulically connected to the Virgin River.
Figure 3. Trilinear diagram showing major-ion chemistry of groundwater and surface water in the Virgin River Gorge subarea, Arizona, Nevada, and Utah.
Download figure to PowerPoint
Cl−:Br− mass ratios can help differentiate different source waters (Davis et al. 1998). Relatively high ratios (>1000) have been reported for surface water from the Virgin River upstream of the study area, compared with generally low ratios (<300) found in regional groundwater (Heilweil et al. 2006; Heilweil and Marston 2011). Cl−:Br− ratios of Littlefield Springs range from 590 to 720 (Table S2), with an average value of 650 (n = 13). It is assumed that a representative end member Cl−:Br− ratio for the carbonate aquifer is about 160, the value for LST (this report). [Correction added after online publication February 21, 2013: Reference to Schaefer et al. 2006 has been removed from the preceding sentence.] An average Virgin River end-member ratio of 890 is based on three samples either from or influenced by the Virgin River (VR-8, VR-11, CPW). Binary mixing calculations using these end-member Cl−:Br− ratios indicate a 68:32 mixture of seepage from the Virgin River and groundwater from the carbonate aquifer.
87Sr/86Sr and Cl−:F− ratios are also useful for evaluating groundwater flow paths. Measured 87Sr/86Sr ratios for Littlefield Springs (LFS-1 through LFS-5) fall within a narrow range (0.7091 to 0.7092; Table 1) and indicate flow through marine carbonate formations. Because carbonate rocks contain very little Rb, and because the maximum 87Sr/86Sr ratio in sea water is 0.709, unaltered marine carbonates typically have ratios of less than 0.710; in contrast, silicate aquifers such as sandstones typically have higher 87Sr/86Sr ratios (McNutt 2000). Similarly narrow ranges have been documented in groundwater from other marine limestone aquifers, such as 0.7072 to 0.7073 for the Lincolnshire Limestone of Great Britain (Bishop et al. 1994) and 0.7079 to 0.7080 for the distal ends of flow paths in the Floridan aquifer (McNutt 2000). Hem (1985) calculated a global Cl−:F− ratio of 2.7 for carbonate-rock aquifers. Hershey et al. (2010) reported Cl−:F− ratios of > 5 for regional springs in the Great Basin carbonate aquifer and 300 to 330 for Littlefield Springs. These higher values than the global mean for carbonate aquifers indicate less dissolution of F−-bearing minerals. Cl−:F− ratios of Littlefield Springs sampled during the current study ranged from 290 to 530 (Table S2), with a mean value of 400 (n = 15). A binary mixture of Virgin River water upstream of Littlefield Springs having a mean Cl−:F− ratio of 640 (n = 10) with carbonate aquifer water having a Cl−:F− ratio of 120 (LST) indicates Littlefield Springs contain a 56:44 mixture of seepage from the Virgin River and groundwater from the carbonate aquifer.
Table 1. Isotopic Concentrations and 14C Ages of Groundwater and Surface Water from the Virgin River Gorge Subarea, Arizona, Nevada, and Utah
|Site||Date||δ18O (‰)||δ2H (‰)||3H (TU)||87Sr/86Sr||δ13C (‰)||14C (pmc)||14C Age (Years)a||14C Age (Years)b|
|LSF-1||March 11, 2009||−12.6||−94.5||0.90 ± 0.10||—||—||—||—||—|
|LFS-1||June 25, 2009||−12.6||−94.7||0.72 ± 0.10||0.7092||−2.97||8.4 ± 0.1||9300||8400|
|LFS-2||June 24, 2009||−12.6||−94.8||0.70 ± 0.10||0.7092||−2.39||9.7 ± 0.1||7000||6000|
|LSF-3||June 23, 2009||−12.6||−94.6||0.93 ± 0.10||0.7091||−3.09||9.4 ± 0.1||8600||7700|
|LFS-4||June 23, 2009||−12.7||−94.2||0.80 ± 0.10||0.7091||−3.12||9.0 ± 0.1||9000||8100|
|LFS-5||June 24, 2009||−12.4||−92.3||1.2 ± 0.10||0.7091||−3.96||20 ± 0.2||3600||3100|
|LFS-6||June 19, 2012||−12.4||−92.3||1.1 ± 0.07||—||−3.48||18 ± 0.1||3900||3200|
|LFS-6 replicate||June 19, 2012||−12.4||−92.3||0.98 ± 0.08||—||−3.48||18 ± 0.1||3900||3200|
|LFS-7||June 20, 2012||−12.4||−93.5||0.76 ± 0.13||—||−3.67||15 ± 0.1||5700||5000|
|LFS-8||June 20, 2012||−12.5||−94.6||0.60 ± 0.06||—||−0.66||10 ± 0.1||1800||1000|
|LFS-9||June 19, 2012||−12.6||−94.9||0.42 ± 0.07||—||−2.88||10 ± 0.1||7700||6700|
|LFS-10||June 20, 2012||−12.7||−95.2||0.49 ± 0.04||—||−2.55||7.6 ± 0.1||9300||8300|
|LFS-11||June 20, 2012||−12.6||−95.1||0.65 ± 0.06||—||−2.96||9.4 ± 0.1||8300||7500|
|LFS-12||June 20, 2012||−12.6||−95.0||0.60 ± 0.09||—||−3.29||12 ± 0.1||6900||6100|
|LFS-13||June 20, 2012||−12.5||−94.8||1.0 ± 0.07||—||−3.47||14 ± 0.1||5900||5300|
|LFS-Palm||June 2, 2010||−12.6||−94.4||—||—||−1.78||14 ± 0.1||2400||1600|
|LST||February 3, 2010||−12.8||−92.8||0.01 ± 0.1||—||−13.2||86 ± 0.2||Modern||1900|
|CPW||February 4, 2010||−12.3||−92.3||2.3 ± 0.10||—||−6.43||63 ± 0.2||Modern||Modern|
|VR-8||March 11, 2009||−12.1||−92.1||2.0 ± 0.10||0.7089||—||—||—||—|
|VR-11||February 4, 2010||—||—||2.0 ± 0.10||—||−4.74||65 ± 0.2||Modern||Modern|
|VR-11||July 21, 2010||−10.7||−85.5||—||—||−6.40||83 ± 0.3||Modern||Modern|
|VR-DS||June 20, 2010||−12.4||−94.0||0.83 ± 0.09|| ||−1.07||12 ± 0.1||1800||640|
δ2H values for Littlefield Springs and VR-DS all have similar values of −92.3 to −94.8 ‰ (Table 1), lying mid-way between and likely comprised of a mixture of Virgin River and high-altitude carbonate aquifer water (Figure 4). While all the waters samples within the study area plot between the global meteoric water line (GML; Dansgaard 1964) and the local Arizona meteoric water line (Kendall and Coplen 2001), both the Virgin River and Littlefield Springs samples show some evaporative enrichment. Hershey et al. (2010) noted that the Littlefield Springs isotopic shift off the GML was larger than for other regional springs in the Great Basin carbonate aquifer. We propose that this larger shift is caused by mixing with evaporatively enriched Virgin River water. Similar to previous Great Basin carbonate aquifer studies by Thomas et al. (2001) and Lundmark et al. (2007), a deuterium mass-balance method was utilized to evaluate these mixing ratios. Assuming a mean δ2H value of −91 ‰ (n = 9) for samples collected along the losing reach of the Virgin River in the Virgin River Gorge (Table 1; Metcalf 1995) and a mean δ2H value of −100 ‰ for four springs at elevations between 2000 and 2200 m located north of the study area (Appendix S1), the mean δ2H value of about −94 ‰ (n = 14) for Littlefield Springs indicates a 67:33 mixture of seepage from the Virgin River and groundwater from the carbonate aquifer.
Figure 4. Relation between δ18O and δ2H values in groundwater and surface water in the Virgin River Gorge subarea, Arizona, Nevada, and Utah.
Download figure to PowerPoint
Like stable isotopes, noble-gas concentrations also provide insight into the recharge conditions of water emanating from Littlefield Springs. Both low total dissolved-gas pressures (about 0.9 to 1.0 atm; Table S1) and Δ20Ne values of −3 to −22 % for most of the Littlefield Springs samples suggest some degassing subsequent to recharge (Aeshbach-Hertig et al. 2008). Acceptable χ2 fits of <3.8 between measured 40Ar, 84Kr, 20Ne, and 129Xe concentrations and CE model calculations for all of the Littlefield Springs sites, however, indicate minimal gas stripping (Table 2). The average noble-gas recharge temperature (Trave) for Littlefield Springs is about 16 °C, which is slightly cooler than the 3-year mean Virgin River temperature of 17 °C measured at VR-8 and consistent with a mixture of waters from the Virgin River and high-altitude recharge in the carbonate aquifer. This mean value excludes LFS-2 because dissolved-gas re-equilibration likely occurred at this spring (Appendix S1).
Table 2. Noble Gases in Groundwater in the Virgin River Gorge Subarea, Arizona, Nevada, and Utah
|Site||Date||Noble Gases||CE Model Results|
|4He (cm3 STP/g)||R/Ra||20Ne (cm3 STP/g)||40Ar (cm3 STP/g)||84Kr (cm3 STP/g)||129Xe (cm3 STP/g)||Δ20Ne (%)||χ2||F||A||Trmax Using Site Altitudea (°C)||Trmin Using Maximum Recharge Altitudeb (°C)||Trave Using Median Altitudec (°C)||Calculated 4HeNUC (cm3 STP/g)|
|LFS-1||June 25, 2009||1.6E−06||0.28||1.42E−07||2.73E−04||3.55E−08||2.29E−09||−9.7||0.01||1.13||0.040||19.5||15.4||17.5||1.5E−06|
|LFS-2||June 24, 2009||1.0E−07||0.46||1.44E−07||2.50E−04||3.09E−08||2.05E−09||−4.8||1.13||1.06||0.117||25.0||20.3||23.2||6.1E−08|
|LFS-3||June 23, 2009||2.3E−05||0.09||1.23E−07||2.67E−04||3.43E−08||2.40E−09||−22.0||1.02||1.52||0.011||18.3||14.0||16.3||2.3E−05|
|LFS-4||June 23, 2009||2.5E−05||0.09||1.40E−07||2.98E−04||3.09E−08||2.64E−09||−13.8||0.00||1.31||0.010||15.3d||11.0d||13.3d||2.5E−05|
|LFS-5||June 24, 2009||5.9E−06||0.08||1.48E−07||2.77E−04||3.82E−08||2.55E−09||−7.5||2.39||1.10||0.116||16.7||13.4||15.6||5.9E−06|
|LFS-6||June 19, 2012||9.8E−06||0.09||1.41E−07||2.99E−04||4.03E−08||2.42E−09||−11.5||1.80||0.00||0.000||17.0||12.0||14.4||9.7E−06|
|LFS-6 repl.||June 19, 2012||1.0E−05||0.09||1.50E−07||3.16E−04||4.12E−08||2.48E−09||−6.9||2.59||0.00||0.000||16.0||10.0||13.4||1.0E−05|
|LFS-7||June 20, 2012||2.3E−05||0.09||1.47E−07||3.14E−04||4.03E−08||2.64E−09||−9.0||0.76||0.00||0.000||15.7||11.3||13.1||2.3E−05|
|LFS-8||June 20, 2012||3.6E−07||0.18||1.52E−07||2.88E−04||3.59E−08||2.36E−09||−3.5||0.29||1.04||0.500||19.1||15.6||17.7||3.2E−07|
|LFS-9||June 19, 2012||2.4E−06||0.10||1.44E−07||2.92E−04||3.74E−08||2.44E−09||−8.9||0.08||0.00||0.000||18.5||13.8||15.7||2.3E−06|
|LFS-10||June 20, 2012||6.7E−07||0.14||1.58E−07||2.66E−04||3.19E−08||1.94E−09||7.5||0.01||0.86||0.078||27.8||23.6||25.3||6.3E−07|
|LFS-11||June 20, 2012||4.4E−06||0.09||1.43E−07||2.89E−04||3.79E−08||2.31E−09||−9.5||0.88||0.00||0.000||19.1||14.4||16.3||4.4E−06|
|LFS-12||June 20, 2012||1.3E−05||0.09||1.44E−07||3.03E−04||3.77E−08||2.61E−09||−10.3||1.30||2.00||0.000||16.8||11.9||14.2||1.3E−05|
|LFS-13||June 20, 2012||2.7E−06||0.26||1.54E−07||3.04E−04||4.07E−08||2.38E−09||−3.6||2.50||0.00||0.000||17.4||14.7||15.7||2.1E−06|
|LST||February 3, 2010||4.6E−08||1.01||1.70E−07||3.18E−04||4.11E−08||2.69E−09||11.8||0.00||0.62||0.000||13.3||11.7||12.5||1.5E−09|
|CPW||February 4, 2010||1.8E−06||0.12||1.73E−07||3.20E−04||3.88E−08||2.73E−09||9.0||1.63||0.00||0.000||16.1||11.3||13.7||1.8E−06|
14C activities of 7.6 to 86 pmC (Table 1) indicate a large range in groundwater age within the Virgin River Gorge subarea. The lowest activities (7.6 to 20 pmC) were measured in Littlefield Springs and VR-DS, consistent with values of 7.8 and 8.9 pmC reported by Hershey et al. (2010). In contrast, samples from CPW, LST, and VR-11 ranged from 63 to 86 pmC. Using measured δ13Caq, an estimated soil-gas δ13C value of −15.8 ‰, and an assumed δ13C value for aquifer materials of 1.5 ‰, the Ingerson and Pearson and the Fontes and Garnier models both yielded plausible ages of 0 and 1900 years, respectively, for LST. This is consistent with other data from the site indicating pre-1950s but late Holocene recharge: no detectable 3H (0.0 TU), low 4HeNUC (1.5 × 10−9 cm3 STP/g), and a high R/Ra value (1.01). Using these same models, the corrected 14C ages of Littlefield Springs are 1 to 9 ka. Assuming these are mixed waters, the carbonate aquifer component is likely much older.
Large 4HeNUC concentrations and low 3He:4He ratios similarly indicate that Littlefield Springs contain old groundwater. 4HeNUC of up to 2.5 × 10−5 cm3 STP/g (Table 2) are higher than previously published values for the Great Basin regional carbonate aquifer. This is supported by R/Ra values for 3He:4He as low as 0.09. A well in the carbonate aquifer of Rush Valley (western Utah) with 4HeNUC of 1 × 10−6 cm3 STP/g and 14C activity of 2.2 pmC yielded an apparent age of 20 to 30 ka (Gardner and Kirby 2011). Wells in the carbonate aquifer in the Snake Valley area of Utah and Nevada have 4HeNUC of 10−6 to 10−7 cm3 STP/g, 14C activities < 4 pmC, and corrected 14C ages >10 ka (Phillip Gardner, U.S. Geological Survey, personal communication, 2011). Thomas et al. (2003) reported 4HeTOT values for carbonate-aquifer springs in southern Nevada up to 1.8 × 10−6 cm3 STP/g and 14C activities of 2.2 to 3.0 pmC. The combination of high 4HeNUC, low 14C activities, and likely dilution with young river water suggests that the carbonate aquifer component of Littlefield Springs may be older than 10 ka.
CFC concentrations show that Littlefield Springs also contain some modern groundwater. Measured CFC and SF6 concentrations in groundwater and Virgin River samples were converted to equivalent atmospheric concentrations (Table 3) based on Henry's Law solubility using the interpreted noble-gas data, including recharge temperature and pressure, recharge salinity, and excess air. Recharge temperatures for the groundwater sites used in the conversion to equivalent atmospheric concentrations ranged from 13 to 25 °C (based on the Trave values given in Table 2). Estimated recharge pressure was determined from altitude/barometric relations using median estimated recharge altitude (average of Hmeas and Hmax) for each site. For the groundwater sites, these altitudes ranged from 1220 to 1580 m above sea level; the land-surface altitude of 650 m was used for VR-11. Estimated recharge salinities ranged from 2000 to 3000 mg/L and were based on binary mixtures of two recharge components: Virgin River seepage loss and carbonate aquifer groundwater (represented by LST). Groundwater excess-air values of 0.0 were based on the unfractionated model (UF) noble-gas concentrations. Figure 5 shows the relation between equivalent atmospheric concentrations of CFC-112 and CFC-13 for water samples collected within the study area. The solid black piston-flow line represents the hypothetical case where groundwater in equilibrium with atmospheric CFCs moves along one flowpath from the recharge area to the point of discharge with minimal longitudinal dispersion and mixing (Zuber 1986; Cook and Böhlke 2000). The colored lines represent binary mixtures of waters from two distinct recharge areas; their points of intersection with the piston-flow line indicates expected CFC concentrations of water that equilibrated with air during particular years (e.g., 1980, 1983, 1985, etc.). Based on their CFC-12 and CFC-113 concentrations, Littlefield Springs may be described with a lumped parameter binary piston-flow model (Eberts et al. 2012) in which young (1980s) recharge is mixed with pre-modern CFC-free groundwater (pre-1950s recharge). The component of modern water in this mix ranges from 31 to 89% (Table 3, excluding LFS-2), with a mean value of 57 ± 16% (1σ). This was determined by dividing the equivalent atmospheric CFC concentration by the measured atmospheric CFC concentration for the estimated recharge year (binary mixing lines in Figure 5). Assuming the modern component is seepage from the Virgin River, groundwater travel times range from 24 to 29 years (26 ± 1.6 years), indicating mid-1980s recharge. CFC concentrations from VR-11 indicate 100% recharge in 2010, as would be expected with water in equilibrium with the atmosphere.
Table 3. Chlorofluorocarbon and Sulfur Hexafluoride Concentrations and Age Estimates for Samples Collected from the Virgin River Gorge Subarea, Arizona, Nevada, and Utah
|Site||Date||CFC-11 Water Conc. (pmol/kg)||CFC-12 Water Conc. (pmol/kg)||CFC-113 Water Conc. (pmol/kg)||CFC-11 Atm Conc. (pptv)||CFC-12 Atm Conc. (pptv)||CFC-113 Atm Conc. (pptv)||CFC-12/ CFC-113 Binary Mix: Modern Component Recharge Year||CFC-12/ CFC-113 Binary Mix: Modern Component Age (Years)||CFC-12 Percent of Modern Component||CFC-113 Percent of Modern Component||SF6 Water Conc. (fmol/kg)||SF6 Atm Conc. (pptv)||SF6 Binary Mix: Modern Component Recharge Year|
|LFS-1||June 25, 2009||0.702||0.587||0.046||58.5||181||12.9||1980||29||59||56||1.68||6.75||Excess|
|LFS-2||June 24, 2009||1.38||1.19||0.156||134||419||52.3||1987||22||98a||97a||1.57||7.58||Excess|
|LFS-3||June 23, 2009||0.582||0.489||0.054||43.1||136||14.4||1985||24||36||34||1.88||7.21||Excess|
|LFS-4||June 23, 2009||0.521||0.472||0.052||33.5||114||10.9||1984||26||31||31||1.79||6.05||Excess|
|LFS-5||June 24, 2009||0.773||1.09||0.099||59.3||280||22.8||1981||28||89||88||2.23||8.21||Excess|
|LFS-6||January 12, 2012||1.38||1.10||0.130||96.9||290||30.3||1985||27||78||75||—||—||—|
|LFS-7||January 12, 2012||1.01||0.81||0.101||66.5||201||21.9||1986||27||56||54||—||—||—|
|LFS-8||January 12, 2012||1.40||1.00||0.144||102b||271b||34.8b||1988||25||61||61||—||—||—|
|LFS-9||January 13, 2012||0.929||0.964||0.134||67.4||262||32.3||1987||25||61||60||—||—||—|
|LFS-10||January 13, 2012||1.08||0.976||0.133||78.1b||265b||32.0b||1988||24||59||53||—||—||—|
|LFS-11||January 13, 2012||0.939||0.820||0.109||68.1||223||26.3||1986||26||55||55||—||—||—|
|LFS-12||June 20, 2012||0.734||0.569||0.072||56.5||163||18.4||1985||27||43||43||—||—||—|
|LFS-13||June 20, 2012||1.21||0.893||0.114||93.2||256||29.2||1986||26||61||62||—||—||—|
|CPW||February 4, 2010||1.24||1.56||0.215||87.3||412||49.9||1987||23||95||92||0.903||2.87||1995|
|LST||February 3, 2010||0.841||1.96||0.232||55.6||487||49.9||1990||20||100||68c||2.30||6.76||2010|
|VR-11||February 4, 2010||2.93||2.71||0.445||149||536||73.4||2010||0||100||97||—||—||—|
Figure 5. Comparison of CFC-12 and CFC-113 equivalent atmospheric concentrations for groundwater and surface water in the Virgin River Gorge subarea, Arizona, Nevada, and Utah.
Download figure to PowerPoint
Alternatives to a Binary Mixing Model
The simplest conceptualization of the mixed groundwater emanating from Littlefield Springs is a lumped parameter binary piston-flow model (BPP). Alternatively, seepage loss along the losing reach of the Virgin River above Littlefield Springs may be described as exponential mixing (E), defined as an infinite number of coalescing groundwater flow paths having an exponential age distribution and typified by uniform recharge to an unconfined aquifer (Cook and Böhlke 2000). The location of the Littlefield Springs samples between the exponential and piston-flow curves on the CFC-12/CFC-113 plot (Figure 5) indicates that in addition to the BPP, these samples could be consistent with BEP (binary model with piston-flow and exponential components), or DM (dispersion models). Such lumped parameter models have been previously invoked when interpreting large multiple-tracer data sets (Eberts et al. 2012). Exponential mixing and then piston flow (EP) would best represent recharge from Virgin River seepage loss if this recharge were relatively evenly distributed. EP has an initial exponential age distribution similar to an exponential mixing model, but is then shifted in time as this assemblage moves toward the discharge location under piston-flow conditions without receiving additional recharge (Cook and Böhlke 2000; Böhlke 2006). Recent seepage studies indicate, however, that almost half of this seepage loss is focused along a short (6-km) reach in the Virgin River Gorge containing known fault zones, including the Grand Wash Fault and Sullivans Canyon Fault (Figure S2). Also, an exponential age distribution would result in a trend of decreasing 3H concentrations (longer and older flow paths) with distance downstream along the 10-km reach of the spring complex, assuming aquifer homogeneity and uniform seepage loss from the Virgin River. In contrast, the similarity in measured 3H concentrations in water from LFS-1 through LFS-13 (0.8 ± 0.2 TU) suggests piston flow from a focused point source rather than an exponential mixture of a large range of flow paths and travel times from the river. An EP model may also describe the older carbonate aquifer component. Exponential mixing could be occurring with in the recharge zone in the northern part of the study area (Figure 1), followed by piston flow as this water moves towards Littlefield Springs. Because this groundwater is likely much older than 60 years and would be CFC-free, an EP mixing model for the carbonate aquifer component could not be evaluated with environmental tracers collected during this study.
Evaluating Virgin River Tritium Content
Virgin River samples VR-8 and VR-11 had 3H concentrations of about 2.0 TU (Table 1). These are lower than current concentrations in precipitation of about 6 TU (Figure S4). The 3H concentrations of Littlefield Springs (0.8 ± 0.2 TU; n = 14) indicate some modern recharge, but older than the Virgin River. The other Virgin River sample, VR-DS, had a lower 3H concentration than VR-8 or VR-11 because it was collected when there was no flow in the river above Littlefield Springs. The estimated 60:40 binary mixture of Virgin River and carbonate aquifer water derived from other environmental tracers (Table 4) and a mean travel time of 26 years based on CFCs (Table 3) can be used for evaluating the initial tritium concentration of seepage from the Virgin River. It is likely that the carbonate-aquifer component of flow discharging from Littlefield Springs is tritium-free based on the long groundwater flow path from the recharge area in the northern part of the study area to Littlefield Springs (Figure 1). Using the mean 3H concentration of Littlefield Springs (0.8 TU), the 60:40 mixture implies that the decay-corrected 3H concentration of Virgin River seepage prior to mixing with the older groundwater was about 1.4 TU. Assuming seepage occurred around 1985, followed by 26 years of radioactive decay (estimated travel time of Virgin River seepage reaching Littlefield Springs), the initial 3H concentration in the losing reach of the Virgin River would have been about 6 TU. This is about one-third of the estimated atmospheric tritium content of 20 TU for 1985 (Figure 2). Virgin River 3H concentrations reported by Trudeau et al. (1983) between 1976 and 1977 have a mean value of 17 TU (excluding those with uncertainty exceeding 30%). Comparing this mean value with the estimated atmospheric tritium content of 52 TU for 1976–1977 yields the same ratio of one-third. This indicates a predominantly tritium-free groundwater source to the Virgin River upstream of the study area, consistent with the discharge of large quantities of old groundwater (14C ages of 1 to 4 ka) from springs along the Virgin River in Zion National Park (Heilweil and Freethey 1992; Kimball and Christensen 1996). This predominantly old Virgin River water is further supported by a comparison of 3H and CFC-12 equivalent atmospheric concentrations (Figure 6), showing that water discharging from Littlefield Springs contains less 3H than expected based on CFC-12 concentrations. Assuming a binary mix of pre-modern carbonate aquifer water with Virgin River seepage loss having a mid-1980s apparent age, the samples would plot between the 1980 and 1990 mixing lines if the Virgin River contained only modern precipitation. VR-11 and CPW (containing seepage loss from the adjacent Virgin River) also have lower 3H concentrations than expected based on CFC-12 concentrations. In contrast to dissolved gases such as CFCs, 3H is a conservative tracer when in contact with the atmosphere. While the low 3H concentrations show that the river water is primarily derived from upstream discharge of old groundwater, higher CFC concentrations are caused by atmospheric re-equilibration of dissolved gases in the river.
Table 4. Estimated Percent of Virgin River Seepage in Littlefield Springs Discharge
|Analysis||Percent from the Virgin River Seepage|
|CFC-12:CFC-113||57 ± 16 (average of 11 values)|
|Average||60 ± 7|
Figure 6. Comparison of 3H and CFC-12 equivalent atmospheric concentrations for groundwater and surface water in the Virgin River Gorge subarea, Arizona, Nevada, and Utah. 3H values for the piston-flow line are atmospheric tritium concentrations decay-corrected to 2010.
Download figure to PowerPoint