Quantifying groundwater discharge from the Valley-Fill Aquifer in Moab-Spanish Valley near Moab, Utah

Moab City and Grand County rely on groundwater for public water supply. Recent development and an increase in water right applications prompted area water managers to call for an updated evaluation of local groundwater resources. The purpose of this study is to (1) prepare a conceptual groundwater flow model for lower Moab-Spanish Valley by delineating flow paths and identifying sources of recharge to the valley-fill aquifer, in order to (2) quantify groundwater outflow to the Colorado River to improve estimates of groundwater available for public use. Samples were collected from 30 wells to analyze major ions, tritium, noble gases, CFCs, SF6, and deuterium and oxygen-18 stable isotopes. The groundwater budget was evaluated by estimating discharge to the Colorado River and loss to Mill Creek. Groundwater discharge was estimated first by performing a Darcy Flux calculation. Twelve new observation wells were drilled and installed in a transect across the Scott M. Matheson Wetlands Preserve, ranging in depth from 25 to 60 feet below ground surface. Eight single-well tests and two dual-well tests were performed to determine transmissivity, which ranged from 90 to 5,400 ft2/day, with a median of approximately 1000 ft2/day. The hydraulic gradient was determined by creating a potentiometric surface map using water levels from both new observation wells and previously existing private wells. Discharge was estimated to be 300 acre-feet per year. A second, independent estimate of groundwater discharge was made using environmental tracer data to determine change in age across some distance along a flow path. 3H/3He ages in the valley-fill aquifer range from 0 to 57 years. Average discharge had a value of 1,000 acre-feet per year. A bromide tracer test was performed to evaluate whether some groundwater was lost to Mill Creek before discharging into the Colorado River. Gain in Mill Creek was found to be negligible. Geochemical properties of valley wells indicate that the valley-fill aquifer is not recharged by water from Glen Canyon Group Aquifer (GCGA), as previously hypothesized by Sumsion (1971); rather, it is more likely recharged by loss from Mill and Pack Creeks.

QUANTIFYING GROUNDWATER DISCHARGE FROM THE VALLEY-FILL AQUIFER IN MOAB-SPANISH VALLEY NEAR MOAB, UTAH by Nora Claire Nelson A thesis submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Master of Science in Geology Department of Geology and Geophysics The University of Utah December 2017 Copyright © Nora Claire Nelson 2017 All Rights Reserved The University of Utah Graduate School STATEMENT OF THESIS APPROVAL The thesis of Nora Claire Nelson has been approved by the following supervisory committee members: D. Kip Solomon , Chair 9/19/2017 Date Approved Victor Heilweil , Member 9/19/2017 Date Approved John M. Bartley , Member 9/20/2017 Date Approved and by Thure Cerling the Department/College/School of and by David B. Kieda, Dean of The Graduate School. , Chair/Dean of Geology and Geophysics ABSTRACT Moab City and Grand County rely on groundwater for public water supply. Recent development and an increase in water right applications prompted area water managers to call for an updated evaluation of local groundwater resources. The purpose of this study is to (1) prepare a conceptual groundwater flow model for lower Moab-Spanish Valley by delineating flow paths and identifying sources of recharge to the valley-fill aquifer, in order to (2) quantify groundwater outflow to the Colorado River to improve estimates of groundwater available for public use. Samples were collected from 30 wells to analyze major ions, tritium, noble gases, CFCs, SF6, and deuterium and oxygen-18 stable isotopes. The groundwater budget was evaluated by estimating discharge to the Colorado River and loss to Mill Creek. Groundwater discharge was estimated first by performing a Darcy Flux calculation. Twelve new observation wells were drilled and installed in a transect across the Scott M. Matheson Wetlands Preserve, ranging in depth from 25 to 60 feet below ground surface. Eight single-well tests and two dual-well tests were performed to determine transmissivity, which ranged from 90 to 5,400 ft2/day, with a median of approximately 1000 ft2/day. The hydraulic gradient was determined by creating a potentiometric surface map using water levels from both new observation wells and previously existing private wells. Discharge was estimated to be 300 acre-feet per year. A second, independent estimate of groundwater discharge was made using environmental tracer data to determine change in age across some distance along a flow path. 3H/3He ages in the valley-fill aquifer range from 0 to 57 years. Average discharge had a value of 1,000 acre-feet per year. A bromide tracer test was performed to evaluate whether some groundwater was lost to Mill Creek before discharging into the Colorado River. Gain in Mill Creek was found to be negligible. Geochemical properties of valley wells indicate that the valley-fill aquifer is not recharged by water from Glen Canyon Group Aquifer (GCGA), as previously hypothesized by Sumsion (1971); rather, it is more likely recharged by loss from Mill and Pack Creeks. iv TABLE OF CONTENTS ABSTRACT ................................................................................................................................... iii LIST OF FIGURES ........................................................................................................................ vi LIST OF TABLES ....................................................................................................................... viii 1 INTRODUCTION ........................................................................................................................ 1 1.1 Background ......................................................................................................................... 1 1.2 Purpose and Scope .............................................................................................................. 2 2 SETTING ..................................................................................................................................... 5 2.1 Geology ............................................................................................................................... 5 2.2 Hydrogeology ..................................................................................................................... 6 3 METHODS ................................................................................................................................... 9 3.1 Drilling and Well Installation ............................................................................................. 9 3.2 Aquifer Testing ................................................................................................................. 10 3.3 Water Level Inventory ...................................................................................................... 12 3.4 Sample Collection and Analysis ....................................................................................... 13 3.5 Bromide Tracer Test ......................................................................................................... 21 4 RESULTS................................................................................................................................... 27 4.1 Aquifer Properties ............................................................................................................. 27 4.2 Hydrochemistry................................................................................................................. 28 4.3 Environmental Tracers ...................................................................................................... 29 4.4 Mill Creek Seepage (Bromide Tracer Test) ...................................................................... 33 5 DISCUSSION ............................................................................................................................ 61 5.1 Conceptual Groundwater Model ....................................................................................... 61 5.2 Groundwater Discharge to Colorado River ...................................................................... 62 6 CONCLUSION .......................................................................................................................... 67 Appendices A: BROMIDE TRACER TEST ................................................................................................... 68 B: AQUIFER TESTING .............................................................................................................. 73 REFERENCES .............................................................................................................................. 78 LIST OF FIGURES Figure 1. Map of the study area. ................................................................................................................ 4 2. Geology of the study area ........................................................................................................... 8 3. Map of sampling network in the study area ............................................................................. 23 4. Preliminary results from an electrical resistivity survey .......................................................... 24 5. Lithologic logs for observation wells installed during this study. ............................................ 25 6. Reference diagram showing air-mixing curves for SF6; CFC-11, CFC-12, and CFC-113; and tritium in precipitation ................................................................................................................... 26 7. Aquifer testing transmissivity results from wetland preserve monitoring wells ...................... 35 8. Potentiometric surface (water table) map ................................................................................. 36 9. Graph displaying specific conductivity (SpC) profiles at U26 and MW-10-D ........................ 37 10. Map of groundwater hydrochemical type............................................................................... 38 11. Map showing the tritium/helium-3 apparent ages from valley-fill aquifer samples............... 39 12. Plot showing the relationship between tritium/helium-3 and sulfur hexafluoride (SF6) apparent ages ................................................................................................................................................ 40 13. Graph showing replicate sulfur hexafluoride (SF6) ................................................................ 41 14. Study area map showing measured sulfur hexafluoride (SF6) concentrations. ...................... 42 15. Plot comparing the calculated sulfur hexafluoride (SF6) partial pressure to total dissolved solids (TDS) .................................................................................................................................. 43 16. Tracer-tracer plots .................................................................................................................. 44 17. Stable isotope results .............................................................................................................. 45 18. Establishing steady state at transport sites.............................................................................. 46 19. Transducer data ...................................................................................................................... 47 20. Synoptic .................................................................................................................................. 48 21. Map depicting a flownet used to calculate the Darcy flux discharge to the Colorado River through the wetland preserve......................................................................................................... 66 22. Flow measurements along Mill Creek that prompted the tracer test ...................................... 71 23. Map of bromide tracer test; location of injection site, transport sites, pre-synoptic, and synoptic ......................................................................................................................................... 71 24. Sample correction (instrumental drift) ................................................................................... 72 vii LIST OF TABLES Table 1. Transmissivity results, square feet per day............................................................................... 49 2. Field Parameters and Alkalinity ............................................................................................... 50 3. Salinity profiles at select sites .................................................................................................. 51 4. Major ion results ....................................................................................................................... 52 5. Measured noble gas concentrations .......................................................................................... 53 6. Closed-system equilibration (CE) model (Aeschbach-Hertig et al., 2000) results, assuming recharge elevation of 1500 m ........................................................................................................ 54 7. Measured and calculated tritium results ................................................................................... 55 8. SF6 results ................................................................................................................................. 56 9. Measured CFC results, pMol/kg............................................................................................... 57 10. Calculated CFC partial pressures ........................................................................................... 58 11. CFC apparent age results ........................................................................................................ 59 12. Stable isotope results, permil .................................................................................................. 60 1 INTRODUCTION 1.1 Background The city of Moab is cradled in the northwest end of Moab-Spanish Valley 1, near the Colorado River, in Grand County. Moab-Spanish Valley extends southeast from Moab into San Juan County, toward the headwaters of Mill Creek in the prominent La Sal Mountains. In 2011, the San Juan Spanish Valley Special Service District (SJSVSSD), which services an unincorporated area in San Juan County, applied to permanently transfer a 5,000 acrefoot per year appropriation for surface water from the San Juan River to wells in Spanish Valley. The proposed transfer was met with opposition from private citizens and local water managers who feared additional groundwater withdrawals could negatively impact existing water rights. In 2013, the Utah Division of Water Rights granted a provisional 600 acre-feet per year out of the 5,000 requested. A decision about the remainder is pending. In 2015, a comprehensive groundwater resource study was designed to help inform this and future water management decisions. The study is jointly funded by the city of Moab, Grand County, San Juan County, the Grand County Water and Sewer Service (GCWSS), the Utah Division of Water Rights (U6i), the Bureau of Land Management (BLM), and the U.S. Forest Service (USFS). The Moab Regional Groundwater Study is led in partnership by the USGS and the University of Utah. The USGS investigated aspects of both recharge and discharge to improve understanding of the aquifer system and its boundaries, and to update and refine the overall groundwater budget for Moab-Spanish Valley and the surrounding area. The USGS study area 1 Moab-Spanish Valley refers to the contiguous topographic feature that, for the purposes of this report, combines the formally distinct political regions, Moab Valley and Spanish Valley, which are separated only by the county line and are located in Grand County and San Juan County, respectively. 2 primarily included the Mill Creek and Pack Creek drainage basins, consisting of Moab-Spanish Valley, the western slopes of the La Sal Mountains, and the slickrock mesas in-between; it extended as far north as Ice Box Canyon and included a small area of Kane Springs Creek to the south (Figure 1). 1.2 Purpose and Scope The purpose of this study is to (1) prepare a conceptual groundwater flow model for lower Moab-Spanish Valley by delineating flow paths and identifying sources of recharge to the valley-fill aquifer, in order to (2) quantify groundwater outflow to the Colorado River to improve estimates of groundwater available for public use. Sumsion (1971) estimated that approximately 8,000 acre-feet of groundwater per year flowed to the Colorado River through the subsurface (not including 3,000 acre-feet consumed by phreatophytes), and concluded that the primary source of recharge to the valley-fill aquifer was premodern Glen Canyon Group Aquifer (GCGA) water from springs and groundwater from the northeast. Gardner (2004), however, found that shallow groundwater near the discharge zone at the Colorado River did not resemble the geochemical signature of GCGA water - the implication being that "unless there is a considerable amount of GCG water discharging from an unknown location, […] the total flow from the GCG aquifer has been significantly overestimated." Gardner (2004) used a Darcy Flux calculation along the length of the valley adjacent to the Colorado River to estimate between 100 and 1,500 acre-ft per year of groundwater discharge to the Colorado River. This study sought to follow up on the findings of that study by collecting additional data to further investigate the groundwater in lower Moab-Spanish Valley, by attempting to locate unaccounted-for GCGA water in order to refine and/or validate the estimate of groundwater discharge to the Colorado River. A conceptual groundwater flow model was developed for lower Moab-Spanish Valley based on geology, physical aquifer properties, and geochemical characteristics. Groundwater 3 samples were collected from 30 wells in lower Moab-Spanish Valley, including 10 new observation wells installed during this study within the Scott M. Matheson Wetlands Preserve (hereafter, the wetland). Samples were analyzed for major ions, dissolved noble gases, tritium, sulfur hexafluoride, chlorofluorocarbons, and stable water isotopes (18O and 2H). Geochemical analyses were used to categorize the samples according to groundwater type, age (time since recharge), recharge elevation, and recharge temperature. With the conceptual model in mind, two independent methods were used to estimate groundwater discharge to the Colorado River. One method used physical aquifer properties to estimate discharge (the Darcy flux method) and the other used geochemical properties (the age gradient method). Lastly, a bromide tracer test was performed on the lower reaches Mill Creek to locate and quantify any groundwater discharging to Mill Creek above the Colorado River. 4 Figure 1. Map of the study area in southeastern Utah, including the city of Moab, Moab-Spanish Valley, the Scott M. Matheson Wetlands Preserve, and the Colorado River. These features comprise the study area (outlined in black), which extends southeast from Moab toward the headwaters of Pack Creek and Mill Creek in the La Sal Mountains. 2 SETTING 2.1 Geology Moab-Spanish Valley is located in the Colorado Plateau physiographic province. It is a northwest-southeast trending topographic feature formed by the collapse of a salt anticline - one of many in the region. The Middle Pennsylvanian Paradox Formation contains sequences of evaporite salts (halite and gypsum), dolomite, and shale, which were deposited in the Paradox Basin, in the shadow of the Uncompahgre Plateau. The buoyant salts migrated into elongated diapirs, under pressure created by deposition of overlying sediments, resulting in both depositionally and tectonically formed anticlines (Doelling, 1983). Groundwater dissolution of the salts resulted in collapse of the anticlines. Paradox Formation caprock contains the leftover anhydrite (dehydrated gypsum) and shale beds after evaporites were leached away; it is exposed in the lower valley along the north and south walls of the canyon, just outside of the wetland (Figure 2). Steep, sandstone walls rim the valley, looming nearly 800 feet above the valley floor. The oldest exposed sandstone in the lower valley is the Triassic Chinle Formation. Overlying the Chinle is the cliff-forming Jurassic Glen Canyon Group, composed of the Wingate, Kayenta, and Navajo Sandstones, in ascending order. The Moab Fault cuts through the center of the valley, downdropping the northeastern block relative to the southwest, though displacement is more pronounced in the valley north of the Colorado River. The prominent La Sal Mountains are igneous in origin. They were intruded in the Paleogene in the form of laccoliths (Doelling et al., 2002), and were subsequently exposed by erosion of overlying sediments. Moab-Spanish Valley is blanketed with Quaternary alluvial sediments deposited by Mill 6 Creek and Pack Creek as well as colluvial sediments from the valley walls. In general, finergrained sediment is deposited toward the lower end of the valley, but paleo-stream channels of Mill Creek and the Colorado River have introduced lenses of coarse gravel. 2.2 Hydrogeology In general, groundwater in Moab-Spanish Valley recharges at high altitudes in the La Sal Mountains, where the tallest peak is just over 12,700 feet above sea level, and discharges into the Colorado River at around 3,950 feet. Moab-Spanish Valley has two major streams, Mill Creek and Pack Creek. Pack Creek joins Mill Creek in downtown Moab before flowing into the Colorado River (Figure 1). There are two aquifers in the study area, the Glen Canyon Group Aquifer (GCGA) and the valleyfill aquifer. The public water supply for Moab City and Grand County is sourced from springs emanating from, and wells completed in, the GCGA. Many irrigation wells produce water from the valley-fill aquifer, which is not a suitable source of culinary water. The GCGA is located in the Glen Canyon Group sandstone formations: the Navajo, the Kayenta, and the Wingate. Although each of these formations contains at its base a lowerpermeability confining bed, fracturing in the northern valley wall along the Moab Fault is sufficient to consider the formations to be hydraulically connected and thus to form one aquifer (Fillmore, 2010). The GCGA is known to contain high-quality water suitable for public water supply and is an EPA-designated sole source aquifer. The GCGA is likely recharged by highelevation precipitation in the La Sal Mountains (Gardner, 2004). The valley-fill aquifer is known to have higher total dissolved solids (TDS) relative to the GCGA, as well as nitrate contamination. A deep brine layer under the wetland is thought to have evolved from groundwater dissolving Paradox Formation salts (Gardner, 2004). The dimensions and extent of the brine are unknown, except where it has been encountered at shallow depths in the wetland. The density 7 gradient between the brine and the overlying fresh groundwater creates a barrier to flow and effectively delineates the bottom of the freshwater aquifer in the wetland. 8 Figure 2. Geology of the study area (modified from Doelling et al., 2002). Moab-Spanish Valley is a northwest-southeast trending topographic feature formed by the collapse of a salt-anticline, related to evaporate salt deposits in the Middle Pennsylvanian Paradox Formation. The geology exposed in the lower valley includes the Triassic Chinle Formation and the Jurassic Glen Canyon Group (Wingate, Kayenta, and Navajo sandstones). The Moab Fault cuts through the center of the valley, downdropping the northeastern block relative to the southwest. Moab-Spanish Valley is blanketed with Quaternary alluvial sediments deposited by Mill Creek and Pack Creek as well as colluvial sediments from the valley walls. The public water supply for the city of Moab comes from the Glen Canyon Group Aquifer, while the valley fill sediment aquifer is used for agricultural watering. 3 METHODS 3.1 Drilling and Well Installation Twelve wells were drilled and installed in the wetland, including eight single-completion and two dual-completion wells (Figure 3). The wells were used for aquifer testing, geochemical sampling, and hydraulic head measurements. Well siting was guided in part by preliminary results from an electrical resistivity survey performed during this study along the Colorado River in the wetland (Briggs, written communication, February 27, 2017). The survey was conducted to locate zones of freshwater discharge along the Colorado River. The data demonstrated, at least qualitatively, that brines are shallow in the north (i.e., the uppermost freshwater layer is thin or nonexistent), whereas a thicker lens of fresh groundwater is probably present along an approximately 1 km stretch from the Mill Creek confluence to the north, and that these are separated by a brackish transition zone in the middle (Figure 4). A Darcy flux calculation requires the dimensions of the cross-sectional area of groundwater flow. Due to the variation in thickness of the fresh groundwater zone revealed by the electrical resistivity survey, the complication of seasonal variation observed by Gardner (2004), and the uncertainty introduced by evapotranspiration in the wetland (Pataki et al., 2005), the transect of wells marking the cross-sectional area was installed on the eastern edge of the wetland rather than along the river (wells U18 through U25, Figure 3). Unfortunately, the central area between U22 and U24 was inaccessible with the drill rig due to the muddy nature of the wetland. The locations of the two well pairs (wells U26 and U27, and wells U28 and U29) were selected to verify the location and depth of the freshwater lens at its thickest point along the Colorado River, as implied by the electrical resistivity survey. 10 Drilling and well installation was performed by RB&G Engineering using a single-axle auger rig. The auger bit was 4 inches in outer diameter with a 1-inch flight, resulting in boreholes that are approximately 6 inches in diameter. Wells ranged in depth from 25 to 61 feet (Figure 5). Each well was constructed with 2.5inch schedule 40 PVC pipe with a 5-foot screened interval above a 1-foot cap. Wells were completed with coarse-grained silica sand around the well screen, bentonite backfill, and 6 feet of cement grout with either a steel or aluminum cap. The materials encountered while drilling were primarily sand and gravel (Figure 5). While gravel did not typically rise to the surface while drilling (it was probably pushed into the sides of the borehole), it was known to be present by shaking and rattling of the drill rig. A splitspoon sample was taken during one such occurrence, and revealed pebbles up to 2 inches in diameter. Wells were completed within high-permeability gravels wherever possible. Wells were developed using a Waterra Inertial Pump operated by a portable actuator until the water was visibly clear prior to aquifer testing with a Grundfos submersible pump. 3.2 Aquifer Testing Eleven aquifer tests were performed on the newly completed observation wells (nine single-well aquifer tests, including one repeat, and two dual-well aquifer tests). A Grundfos submersible pump was used to create drawdown, which was recorded every second on either a Hobo or Troll transducer (1-second data were later reduced to 1-minute data for analysis). Pumping rates were measured throughout the test using a calibrated 5-gallon bucket and stopwatch and ranged from approximately 0.3-5 gpm. Duration of pumping was approximately 3 hours, after which water levels were allowed to recover for at least 30 minutes or until they had returned to static level. (Complete drawdown data for each test are found in section A-2.) Where possible, transmissivity was estimated using the Cooper-Jacob (1946) straight-line method for drawdown data and the Theis (1935) recovery method for recovery data. In addition, 11 transmissivity was estimated from specific capacity. The Cooper-Jacob (1946) straight-line method is a graphical approach to evaluating aquifer properties from drawdown in a well or wells over time. From drawdown data, transmissivity (𝑇𝑇) was calculated as 𝑇𝑇 = 2.3𝑄𝑄 4𝜋𝜋∆𝑠𝑠 (1) where 𝑄𝑄 is the pumping rate, and ∆𝑠𝑠 is the change in drawdown corresponding to one log-cycle of time on a line fit to the late-time data on a semilog plot of time versus drawdown data. Similarly, transmissivity was calculated from recovery data as 𝑇𝑇 = 2.3𝑄𝑄 4𝜋𝜋∆𝑠𝑠′ (2) where ∆𝑠𝑠′ is the change in recovery corresponding to one log-cycle of time on a semilog plot of 𝑡𝑡/𝑡𝑡′ versus ∆𝑠𝑠′, where 𝑡𝑡 is time since pumping started, and 𝑡𝑡′ is time since pumping stopped (Theis, 1935; Brown et al., 1963). The Cooper-Jacob method is an approximation of the Theis solution, and as such, the assumptions remain that the aquifer is fully confined, of infinite extent, and uniform thickness; the well is fully penetrating; and the pumping rate is constant. The Cooper-Jacob approximation 𝑟𝑟 2 𝑆𝑆 is valid for late-time data when pumping duration is sufficiently long, i.e., when 𝑢𝑢 = 4𝑇𝑇𝑇𝑇 < 0.01, where 𝑟𝑟 is the radius of the well for single-well tests and the distance from the pumping well to the observation well or wells in multiple-well tests; 𝑆𝑆 is storativity (approximately equal to specific yield (Sy) in an unconfined aquifer, assumed to be 0.3); 𝑇𝑇 is transmissivity, and 𝑡𝑡 is time since pumping began (Cooper and Jacob, 1946). Values of 𝑢𝑢 ranged from 8E-7 to 0.008, sufficiently small (less than 0.01) to justify the use of the Cooper-Jacob method. 12 Several assumptions were violated - namely, the aquifer was unconfined and the wells were not fully penetrating. However, a study by Halford et al. (2006) that compared transmissivity estimates of single-well tests using Cooper-Jacob analysis to known values found that "more than 90% of the unconfined aquifer transmissivities […] were within a factor of 2 of the known values" and concluded that "interpretation of single-well tests with the Cooper-Jacob method remains more reasonable than most alternatives." Transmissivity was estimated from specific capacity by developing an empirical equation for the area, similar to Driscoll (1986), using the following equation from Theis (1935): 𝑇𝑇 = 𝑄𝑄 𝑊𝑊(𝑢𝑢) 4𝜋𝜋𝑠𝑠𝑤𝑤 = 𝑊𝑊(𝑢𝑢) 𝑄𝑄 4𝜋𝜋 𝑠𝑠𝑤𝑤 (3) where 𝑄𝑄 ⁄𝑠𝑠𝑤𝑤 is the specific capacity of the well (the ratio of the pumping rate to the drawdown). The [𝑊𝑊(𝑢𝑢)/4𝜋𝜋] term was developed from transmissivity data produced from the other methods, resulting in the following empirical relationship: 𝑇𝑇 = 10 𝑄𝑄 𝑠𝑠𝑤𝑤 (4) 3.3 Water Level Inventory Water levels were measured over a 5-day period in February 2016. Each measuring point was surveyed using a Trimble Real Time Kinematics (RTK) GPS to attain the elevation above mean sea level, to an uncertainty of, for the most part, less than 0.1 feet. Water levels were taken from the measuring point with either a chalked steel tape or an electronic water level probe. 13 3.4 Sample Collection and Analysis Groundwater and surface water samples were collected in Spanish Valley and the surrounding area to characterize water types and to delineate groundwater flow paths. The sample data also provided insight into the fate of high-quality Glen Canyon Group Aquifer water and the source(s) of recharge into the Spanish Valley valley-fill aquifer. Prior to sample collection, a minimum of three casing-volumes of water were purged from each well and field parameters were allowed to stabilize. 3.4.1 Field Parameters and Alkalinity Field parameters, including temperature, specific conductance, pH, total dissolved gases (TDG), and dissolved oxygen (DO), were collected on a calibrated Hydrolab multiparameter water quality probe. The probe was also used to take salinity profiles at two locations in the wetland, U26 and MW-10-D, to delineate the freshwater to brine transition with depth. Readings were taken every foot; the water column and probe were allowed to equilibrate for approximately 1 minute each time the probe was moved to a new position. Alkalinity was measured in the field in mg/L as CaCO3 using a Hach digital titrator. 3.4.2 Major Ions Major-ion samples were pumped through a 0.45-micron filter capsule into 250 mL polyethylene bottles that had been triple-rinsed with formation water. Cation and anion samples were collected separately, the former being preserved with nitric acid. Surface water samples were collected directly from the source, filtered, and preserved on-site. Samples were kept refrigerated before analysis. Samples were analyzed on a Metrohm 883 Basic IC Plus ion chromatograph at the Geomicrobiology Laboratory at the University of Utah in Salt Lake City, Utah. 14 3.4.3 Dissolved Noble Gases Concentrations of dissolved noble gases can be used to determine groundwater recharge temperature and elevation. Noble gases in groundwater have three significant sources: (1) exchange with the atmosphere prior to recharge, (2) radioactive isotopes, and (3) excess air (Aeschbach-Hertig et al., 2000). The partitioning of atmospheric gases into liquid water is described by Henry's law, 𝑝𝑝𝑖𝑖 = 𝐻𝐻𝑖𝑖 𝐶𝐶𝑖𝑖 (5) where 𝑝𝑝𝑖𝑖 is the partial pressure of gas i in air, 𝐻𝐻𝑖𝑖 is the Henry solubility constant (empirically determined for each gas, dependent on temperature and salinity), and 𝐶𝐶𝑖𝑖 is the dissolved gas concentration at equilibrium. Recharging water in the unsaturated zone is constantly equilibrating with air until it reaches the water table and is cut off from atmospheric exchange, thus preserving the dissolved gas concentration. If noble gas concentrations in each groundwater sample can be corrected for radiogenic isotopes 2 and excess air, the solubility coefficients can be used to reconstruct the temperature at which the groundwater recharged. Excess air describes the occurrence of dissolved gas concentrations in groundwater that exceed what is possible through solubility equilibrium alone. Excess air often contains gas ratios similar to those in the atmosphere. Its occurs as air becomes trapped in pore spaces during water table fluctuations; the entrapped bubbles may be partially or fully incorporated into the groundwater through dissolution and diffusive gas exchange (Stute and Schlosser, 2000). Dissolved noble gas samples were collected using the copper tube method according to the procedure outlined by the University of Utah Dissolved and Noble Gas Laboratory 2 Radiogenic isotopes of helium, 3He and 4He, are produced by 3H and uranium-thorium (U-Th) decay, respectively. Helium isotopes were not used in this study to calculate recharge temperature or excess air, but they are measured for the purposes of 3H/3He dating, and 4He accumulation is a good qualitative indication of old groundwater. 15 (http://www.noblegaslab.utah.edu/pdfs/cu_tube_sampling.pdf). Samples were collected in 3/8inch diameter copper tubes cut to approximately 30 inches in length. To sample, water was pumped through a copper tube from a connection as close to the wellhead as possible. The outflow was back-pressured with a regulator valve to keep the dissolved gases in solution until the sample could be sealed with refrigeration clamps. Special care was taken to ensure that gases were not introduced to or allowed to escape from the sample during collection. Samples were collected in duplicates. Samples were analyzed at the Dissolved and Noble Gas Laboratory at the University of Utah in Salt Lake City. Prior to analysis, the dissolved gases were extracted from the water sample. In a closed system under high vacuum, the water sample was transferred from the copper tube to a stainless-steel flask. The dissolved gases were driven into a second flask using a temperature gradient induced by heating the water sample in the first flask and chilling the second flask with liquid nitrogen. The gas sample was sealed again until it was transferred to the mass spectrometer. The heavier gases (Ne, Ar, Kr, and Xe) were analyzed on a Stanford Research Systems RGA300 quadrupole mass spectrometer. Helium isotopes (3He and 4He) were analyzed on a Mass Analyzer Products 215-50 magnetic sector field mass spectrometer. Recharge temperature and excess air were determined using the closed-system equilibration (CE) model, which "assumes equilibrium is attained in a closed system of initially air-saturated water and finite volume of entrapped air under constant hydrostatic pressure," 𝐶𝐶𝑖𝑖 = 𝐶𝐶𝑖𝑖∗ + (1−𝐹𝐹)𝐴𝐴𝑒𝑒 𝑧𝑧𝑖𝑖 1+𝐹𝐹𝐴𝐴𝑒𝑒 𝑧𝑧𝑖𝑖 ⁄𝐶𝐶𝑖𝑖∗ where 𝐶𝐶𝑖𝑖 is the gas concentration in solution, 𝐶𝐶𝑖𝑖∗ is the moist-air solubility equilibrium (6) concentration (a function of temperature, salinity, and total atmospheric pressure), 𝐴𝐴𝑒𝑒 is the initial 16 volume ratio of trapped air to water, 𝑧𝑧𝑖𝑖 is volume fraction of each gas in dry air, and F is the fractionation parameter describing the degree of excess air fractionation from no excess air to pure excess air (Aeschbach-Hertig et al., 2000). The system of four equations, one for each of four gases (Ne, Ar, Kr, and Xe), is sufficient to solve for three parameters (𝐴𝐴𝑒𝑒 , recharge temperature, and F). (Salinity was assumed to be negligible because the source of recharge is meteoric; see section 4.3.4 on stable isotopes.) Although in concept it is possible to also solve for recharge elevation (if unknown), the system of equations would no longer be overdetermined and measurement errors would lead to a nonunique determination of elevation (Manning and Solomon, 2003). A best fit model is determined by minimizing the sum of chi-squared (χ2 ), 𝜒𝜒 2 = ∑𝑖𝑖 2 𝐶𝐶𝑖𝑖 −𝐶𝐶𝑖𝑖𝑚𝑚𝑚𝑚𝑚𝑚 𝜎𝜎𝑖𝑖2 (7) where, 𝐶𝐶𝑖𝑖 is the measured concentration of gas i, 𝐶𝐶𝑖𝑖𝑚𝑚𝑚𝑚𝑚𝑚 is the modeled concentration, and 𝜎𝜎𝑖𝑖 is standard deviation in the measurements (Aeschbach-Hertig et al., 1999). The modeled concentrations are generated by perturbing the model parameters (𝐴𝐴𝑒𝑒 , recharge temperature, and F) within a theoretical range. 3.4.4 Tritium Tritium (3H) is a radioactive isotope of hydrogen used to date young groundwater. Tritium decays by beta emission to the noble gas 3He with a half-life of 12.32 years. Small amounts of tritium are generated naturally through cosmic bombardment in the upper atmosphere, but above-ground nuclear testing in the 1950s and 1960s introduced large quantities 17 of tritium to the atmosphere, increasing natural background concentrations of 3-6 TU3 (Kaufman and Libby, 1954) to concentrations in excess of 5000 TU at its peak in the 1960s (Solomon and Cook, 2000) (Figure 6). Tritium samples were collected with no head-space in 500 mL low density polyethylene (LDPE) bottles, triple-rinsed with formation water. A duplicate sample was collected as backup. The 3H/3He age is defined as 𝑡𝑡3𝐻𝐻/ 3𝐻𝐻𝐻𝐻 = 𝜆𝜆−1 𝑙𝑙𝑙𝑙 3𝐻𝐻𝐻𝐻 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 3𝐻𝐻 + 1 (8) where 𝑡𝑡3𝐻𝐻/ 3𝐻𝐻𝐻𝐻 is the 3H/3He age, 𝜆𝜆 is the 3H decay constant, and 3𝐻𝐻𝐻𝐻𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 is tritiogenic 3He (Solomon and Cook, 2000). The 3H component was measured at the Dissolved and Noble Gas Laboratory at the University of Utah using the helium ingrowth method (Clarke et al., 1976) using a Helix Split Flight Tube (SFT) sector field mass spectrometer. The tritiogenic 3He component was attained by correcting total 3He ( 3𝐻𝐻𝐻𝐻𝑡𝑡𝑡𝑡𝑡𝑡 ) (see section 3.4.3 on dissolved noble gas methods for description of helium measurement). The total amount of 3He in the sample can be expressed as 3 𝐻𝐻𝐻𝐻𝑡𝑡𝑡𝑡𝑡𝑡 = 3𝐻𝐻𝐻𝐻𝑎𝑎𝑎𝑎𝑎𝑎 + 3𝐻𝐻𝐻𝐻𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 + 3𝐻𝐻𝐻𝐻𝑛𝑛𝑛𝑛𝑛𝑛 + 3𝐻𝐻𝑒𝑒𝑚𝑚𝑚𝑚𝑚𝑚 (9) where 3𝐻𝐻𝐻𝐻𝑎𝑎𝑎𝑎𝑎𝑎 is helium from the atmosphere, 3𝐻𝐻𝐻𝐻𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 is helium produced by tritium decay used for age dating, 3𝐻𝐻𝐻𝐻𝑛𝑛𝑛𝑛𝑛𝑛 is helium produced by nuclear reactions in the subsurface, and 3𝐻𝐻𝐻𝐻𝑚𝑚𝑚𝑚𝑚𝑚 is helium from the mantle (Solomon and Cook, 2000). Mantle sources of 3He were assumed to be negligible. Atmospheric helium is further subdivided into two components: 3 Tritium concentrations are reported in tritium units (TU), where one TU is equal to one molecule of 3H1HO in 1018 molecules of H2O. 1 18 3 𝐻𝐻𝐻𝐻𝑎𝑎𝑎𝑎𝑎𝑎 = 3𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠𝑠𝑠 + 3𝐻𝐻𝐻𝐻𝑒𝑒𝑒𝑒𝑒𝑒 (10) where 3𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠𝑠𝑠 is from solubility equilibrium with the atmosphere, and 3𝐻𝐻𝐻𝐻𝑒𝑒𝑒𝑒𝑒𝑒 is from excess air (Solomon and Cook, 2000). The following equation from Solomon and Cook (2000) was used to solve for 3𝐻𝐻𝐻𝐻𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 : 3 𝐻𝐻𝐻𝐻𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 = 4𝐻𝐻𝐻𝐻𝑚𝑚 𝑅𝑅0 − 𝑅𝑅𝑠𝑠𝑠𝑠𝑠𝑠 [ 4𝐻𝐻𝐻𝐻𝑠𝑠𝑜𝑜𝑜𝑜 + (𝑁𝑁𝑁𝑁𝑚𝑚 − 𝑁𝑁𝑁𝑁𝑠𝑠𝑠𝑠𝑠𝑠 )𝛼𝛼 ′ 𝑅𝑅𝐻𝐻𝐻𝐻−𝑁𝑁𝑁𝑁 ] −𝑅𝑅𝑟𝑟𝑟𝑟𝑟𝑟 [ 4𝐻𝐻𝐻𝐻𝑚𝑚 − 4𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠𝑠𝑠 − (𝑁𝑁𝑁𝑁𝑚𝑚 − 𝑁𝑁𝑁𝑁𝑠𝑠𝑠𝑠𝑠𝑠 )𝑅𝑅𝐻𝐻𝐻𝐻−𝑁𝑁𝑁𝑁 ] (11) where 4𝐻𝐻𝐻𝐻𝑚𝑚 is total measured 4𝐻𝐻𝐻𝐻, 𝑅𝑅0 is the 3𝐻𝐻𝐻𝐻/ 4𝐻𝐻𝐻𝐻 ratio in the sample at time of collection, 𝑅𝑅𝑠𝑠𝑠𝑠𝑠𝑠 is the 3𝐻𝐻𝐻𝐻/ 4𝐻𝐻𝐻𝐻 expected ratio for water in equilibrium with the atmosphere at the specified recharge elevation, 4𝐻𝐻𝐻𝐻𝑠𝑠𝑠𝑠𝑠𝑠 is the expected 4𝐻𝐻𝐻𝐻 concentration for water in equilibrium with the atmosphere at the specified recharge elevation, 𝑁𝑁𝑁𝑁𝑚𝑚 is total measured neon, 𝑁𝑁𝑁𝑁𝑠𝑠𝑠𝑠𝑠𝑠 is the solubility neon concentration, 𝛼𝛼 ′ is the air-water isotope fractionation factor, 𝑅𝑅𝐻𝐻𝐻𝐻−𝑁𝑁𝑁𝑁 is the ratio of helium to neon in the atmosphere, and 𝑅𝑅𝑟𝑟𝑟𝑟𝑟𝑟 is the ratio of 3𝐻𝐻𝐻𝐻𝑛𝑛𝑛𝑛𝑛𝑛 / 4𝐻𝐻𝐻𝐻𝑟𝑟𝑟𝑟𝑟𝑟 . 3.4.5 Sulfur Hexafluoride SF6 is an industrial compound whose presence in the atmosphere can be used to date young groundwater. Industrial production of SF6 began in 1953 for its use as an electrical insulator (Busenberg and Plummer, 2000). It was first detected in the atmosphere in 1970 at 0.03 pptv (Lovelock, 1971). The low solubility of SF6 in water and subsequent long residence time in the atmosphere has allowed the atmospheric mixing ratio to increase steadily over time to the current (January 2017) value of 9.26 pptv (Figure 6). Its stable, non-reactive nature, even in very reducing environments (Wilson and Mackay, 1993), adds to its usefulness as a tracer. Although most SF6 in groundwater is anthropogenic in origin, it is naturally produced in relatively small quantities in some igneous environments (Koh et al., 2007); Heilweil (2014) also found evidence 19 of natural production of SF6 from crustal sources. SF6 samples are collected in 1-liter amber glass bottles that were safety coated on the outside with plastic, and sealed with Polyseal cone-lined caps. To collect the sample, tubing from the pump was placed at the bottom of the sample bottle, allowing water to overflow until at least three sample-volumes have been purged through the sample bottle. The bottle was then capped with no head space and sealed with electrical tape. Samples were collected in duplicate. After collection, the samples were kept in a cooler to prevent overheating, water expansion, and bottle breakage. SF6 samples were analyzed at the Dissolved and Noble Gas Laboratory at the University of Utah on a Shimadzu GC-8A gas chromatograph. The resulting measured concentrations were corrected for excess air, determined independently from noble gas analysis (see section 3.4.3). The partial pressure of SF6 during air-water equilibrium at the water table prior to recharge was calculated using the following equation from Busenberg and Plummer (2000): 𝑥𝑥𝑆𝑆𝐹𝐹6 = 𝑐𝑐𝑆𝑆𝑆𝑆6 𝐾𝐾𝐻𝐻 𝑃𝑃 − 𝑝𝑝𝐻𝐻2 𝑂𝑂 (12) where 𝑥𝑥𝑆𝑆𝑆𝑆6 is the dry air mole fraction of SF6, 𝑐𝑐𝑆𝑆𝑆𝑆6 is the concentration of SF6 in the sample, corrected for excess air, 𝐾𝐾𝐻𝐻 is the Henry's law constant, 𝑃𝑃 is the total atmospheric pressure, and 𝑝𝑝𝐻𝐻2 𝑂𝑂 is the partial pressure of water. 𝐾𝐾𝐻𝐻 was calculated according to Bullister et al. (2002) and is a function of salinity and recharge temperature (determined from dissolved noble gas analysis). The resulting calculated partial pressure was related back to the air mixing curve through time (Figure 6). Data prior to 1970 have been reconstructed by Maiss et al. (1994); subsequent data were taken from NOAA measurements at Niwot Ridge, Colorado. 20 3.4.6 Chlorofluorocarbons CFCs are industrial compounds whose presence in the atmosphere can be used to date young groundwater. CFCs were developed in the 1930s as refrigerants. Overall production was limited in 1987, after it was discovered that CFCs were significant contributors to atmospheric ozone depletion. Production nearly ceased entirely in 1996 in accordance with the Clean Air Act. Air mixing ratios peaked in 1994, 2002, and 1995, for CFC-11, CFC-12, and CFC-113, respectively (Figure 6). CFC-11, CFC-12, and CFC-113 can be used to date groundwater back to 1947, 1941, and 1955, respectively (Plummer and Busenberg, 2000). CFC samples were collected in 125 mL clear glass bottles, sealed with foil-lined caps. Samples were collected through refrigeration-grade copper tubing to avoid desorption of atmospheric CFCs from plastic tubing. Before collection, sample bottles and caps were triplerinsed with formation water. To collect samples, copper tubing was inserted into the sample bottle to the bottom. Once the bottle is overflowing, it is submerged in a bucket of formation water. The bottle was allowed to overflow until at least three sample-volumes were purged through the bottle, and was capped while still underwater, without head space. Samples were collected per the procedure outlined by the USGS Reston Groundwater Dating Laboratory. Samples were collected in sets of four, and were stored at room temperature away from sunlight before analysis. CFC samples were analyzed at the Dissolved and Noble Gas Laboratory at the University of Utah on a custom-fabricated line using a Shimadzu purge and trap system based on Bullister et al. (2002). 3.4.7 Stable Isotopes Stable isotope (𝛿𝛿 18O and 𝛿𝛿 2H) compositions can be used to trace groundwater provenance because of the physical processes that govern their distribution. Stable isotope compositions are reported in 𝛿𝛿 (delta) notation, 21 𝛿𝛿 = 𝑅𝑅 𝑅𝑅𝑠𝑠𝑠𝑠𝑠𝑠 − 1 × 1000 ‰ (13) where 𝑅𝑅 is the ratio of 18O/16O or 2H/1H in the sample, 𝑅𝑅𝑠𝑠𝑠𝑠𝑠𝑠 is the ratio in the standard (VSMOW, Vienna Standard Mean Ocean Water), and ‰ is the unit permil. Isotopic compositions of meteoric waters are linearly correlated, falling along the global meteoric water line (GMWL) defined by Craig (1961) as: 𝛿𝛿𝛿𝛿 = 8 𝛿𝛿 18𝑂𝑂 + 10 (14) Mass-dependent isotope fractionation during phase changes (i.e., evaporation, condensation) results in enrichment of the heavier isotope in the denser phase (e.g., during precipitation, the heavy isotopes, 18O and 2H, are preferentially rained out). Removal of the precipitated water from the cloud mass (by a process approximately described by Rayleigh distillation) results in increasingly isotopically depleted waters, which can be correlated geographically with latitude, distance from shorelines, and altitude. Stable isotope samples were pumped through a 0.45-micron filter capsule into 60 mL polyethylene bottles that had been triple-rinsed with formation water. Stable isotope samples were analyzed in the Spatio-Temporal Isotope Analytics Laboratory (SPATIAL) on a Picarro CRDS (cavity ring-down spectroscopy) water isotope analyzer. 3.5 Bromide Tracer Test A bromide tracer test was performed along Mill Creek near the Colorado River to evaluate whether groundwater was discharging into Mill Creek before reaching the Colorado River. The need for the tracer test was prompted by flow measurements taken with a SonTek FlowTracker Handheld-ADV (Acoustic Doppler Velocimeter) that indicated a gain of 22 approximately 1 cfs on a 1 mile reach of lower Mill Creek. A bromide tracer injection was designed to locate and quantify the gain. A bromide tracer injection uses a concentrated solution of sodium bromide (NaBr) injected at a constant, known rate, whereby any dilution in measured concentrations of samples taken downstream indicate the occurrence of groundwater inflow (seepage) into the stream. See a detailed description of bromide tracer test methods in section A1. 23 Figure 3. Map of sampling network in the study area, showing both the locations of newly installed observation wells and pre-existing wells. Wells sampled by the University of Utah in 2015-2016 are indicated in red; bright red indicates locations where wells were drilled and installed during the 2015-2016 field effort. Locations sampled by the U.S. Geological Survey (USGS) (data included in this report) are indicated in green; these locations include groundwater wells (circles), springs (diamonds), and surface water (triangles). 24 Figure 4. Preliminary results from an electrical resistivity survey (modified from Briggs, written communication, February 27, 2017) to investigate shallow alluvial groundwater discharge into the Colorado River. The map shows the total electrical conductivity (EC) in milliSiemens (mS) for shallow alluvial aquifer groundwater in the Matheson Wetlands. The data demonstrated groundwater discharge in the north is dominated by shallow brines with thin to no indication of freshwater. However, the data indicate that a thicker lens of fresh groundwater is present along the 0.5-mile reach from the Mill Creek confluence to the north. The brine and freshwater discharge areas are separated by a brackish transition zone in the middle. 25 Figure 5. Lithologic logs for observation wells installed during this study, showing simplified lithology, depth, and well screen interval. The 12 observation wells include 8 single-completion wells (U18 through U25) and two dual-completion wells (U26, U27; and U28, U29). The lithologic logs are aligned from approximately north to south (see Figure 3 for locations). 26 600 2500 CFC Mixing Ratio, pptv 500 CFC12 2000 400 1500 300 1000 200 500 100 0 1940 SF6 Mixing Ratio x 100, pptv; Tritium in precipitation, TU CFC11 0 1950 1960 1970 1980 1990 2000 2010 Figure 6. Reference diagram showing air-mixing curves for SF6 (in pptv x 100); CFC-11, CFC12, and CFC-113 (in pptv); and tritium in precipitation (in tritium units, TU) (USGS Groundwater Dating Laboratory). 4 RESULTS 4.1 Aquifer Properties 4.1.1 Transmissivity Transmissivity estimated using the Cooper-Jacob straight-line method for drawdown data ranged from 60 to 4,100 ft2/day (Table 1). In several cases, drawdown affected the pumping rate in such a way that the water level experienced an initial dramatic drop and then rose steadily for the remainder of the test (see U23 and U24 drawdown curves, section A-2); as a result, in these cases, transmissivity could not be estimated from Cooper-Jacob analysis of the drawdown data. Transmissivity estimated from specific capacity ranged from 80 to 6,200 ft2/day. Transmissivity estimated using the Theis recovery method ranged from 60 to 5,900 ft2/day. The recovery data had too much noise in the case of U18, and the test was terminated prematurely in the case of U23. Therefore, transmissivity could not be estimated in these cases. Overall, average transmissivities at each aquifer test site using all available methods ranged from 90 to 5,400 ft2/day, with a median of approximately 1000 ft2/day (Figure 7). Standard deviation at each test site ranged from 0 to 920 ft2/day. 4.1.2 Potentiometric Surface (Water Table) Hydraulic head values derived from water-level measurements were contoured to evaluate general directions of groundwater flow (Figure 8). Head values range from 4,000 to 3,950 in the lower end of the valley. There is a notable flattening of the hydraulic gradient (from 0.02 to 0.005) to the west / northwest of the 3,990 ft contour. This could signify a change in transmissivity, either because of an increase in aquifer thickness or an increase in hydraulic conductivity. The hydraulic head contours abut the north valley wall at an angle less than 90 28 degrees, implying that some amount of groundwater is moving from the GCGA to the valley-fill aquifer along the north wall of the valley near the wetland. 4.2 Hydrochemistry 4.2.1 Field Parameters Field parameters and alkalinity data are summarized in Table 2. The lowest measured specific conductivity was 680 µS/cm, and two samples exceeded the meter's detection limit of 100,000 µS/cm. The data from specific conductivity profiles at two locations (U26 and U28) are presented in Table 3; the data show the transition to brine occurring at approximately 30 feet below the measuring point in both wells (Figure 9). 4.2.2 Major Ions and Alkalinity Major ion concentrations are summarized in Table 4. Charge balances ranged from 0 to 23%. For freshwater samples, charge balances were within 10%. Samples with poor charge balances (greater than 10%) were either from brine or brine-affected water. (The brines are high in ammonia, and the elution time between chloride and ammonia is small, making it difficult to separate the peaks.) Total dissolved solids (TDS) ranged from 533 to 159,201 mg/L. Stiff diagrams of the major-ion chemistry (Figure 10) demonstrate that water types generally fall into three categories: (1) low-TDS calcium-bicarbonate, (2) moderate-TDS calcium-sulfate, and (3) high-TDS sodium-chloride. Alkalinity ranges from 124 to 314 mg/L as CaCO3. High-quality low-TDS calcium-bicarbonate waters are characteristic of the Glen Canyon Group Aquifer (Steiger and Susong, 1997; Gardner, 2004). This geochemical signature is exhibited by wells located on the Glen Canyon Group slickrock plateau between Spanish Valley and the La Sal Mountains, groundwater emanating from springs on the north wall of Spanish Valley where it abuts the plateau, and surface water in Mill Creek prior to entering Spanish Valley. Moderate-TDS calcium-sulfate type waters are ubiquitous in the valley-fill aquifer, and 29 are also found in Pack Creek before it enters the valley. The high-TDS sodium-chloride brines are located at the distal end of the valley, near the Colorado River, and are attributed to dissolution of Paradox Formation salts. 4.3 Environmental Tracers 4.3.1 Dissolved Noble Gases Measured dissolved noble gas concentrations are presented in Table 5. Values of R/Ra ranged from 0.067 to 2.053. Calculated recharge temperature (Trech) and excess air (Ae) results calculated using the closed-system equilibration (CE) model assuming a recharge elevation of 1500 m are presented in Table 6. Recharge temperatures ranged from 8 to 19ºC (with one outlier of 30 ºC removed), and excess air ranged from 4.3E-4 to 0.15. Most samples showed good fits to the CE model; the average sum of chi squared was 0.5. The recharge temperature and excess air parameters were used in the analyses of 3H/3He, SF6, and CFC apparent ages. 4.3.2 Tritium Measured and calculated parameters associated with 3H/3He age dating are presented in Table 7. Measured tritium values ranged from 0.01 to 4.79 TU. Calculated 4Heterr ranged from 3.47×10-9 to 9.37×10-6. Calculated 3Hetrit ranged from -3.35 to 163 TU, resulting in 3H/3He ages that ranged from 0 § to 164 years. In general, the 3H/3He ages increase down-gradient (Figure 11). The uncertainty associated with 3H/3He ages ranges from 0 to 154 years, with many values falling between 0 and 8 years. The uncertainty is calculated by perturbing the parameters in equation 11 within their respective ranges of error. Groundwater ages of old waters are less sensitive to error in the 3Hetrit component, because the proportion of 3Hetrit to total measured 3He § Negative 3H/3He ages from negative 3Hetrit were reported with an age of 0 years. 30 increases in any given sample over time as more 3H decays to tritiogenic 3He. However, the 3 H/3He age uncertainty can also increase with large values of excess air and terrigenic 4He; the fraction of total He that comes from 3Hetrit decreases with increasing 3HeAe (from excess air) and 3 Herad (from 4Heterr), increasing the uncertainty in 3Hetrit, and with it, the apparent age (Solomon and Cook, 2000). For mixtures of young and old (i.e., tritiated and nontritiated) waters, the 3H/3He age is biased toward the young fraction (Solomon and Cook, 2000). Several samples were flagged as possible mixtures of young and old water. For example, the U11 sample had a calculated age of 0 years, which is consistent with a measured tritium of 3.14 TU (Table 7), but it has a low R/Ra value, which is a first-order indicator of the presence of older water. An R/Ra of less than 1 is likely due to the presence of terrigenic helium-4, which builds up over time in old waters due to radiogenic decay of U and Th (Solomon and Cook, 2000). In addition to analytical error, there are uncertainties due to sampling: ages vary due to the depth of the well, and the screened interval of the well, and the permeability of the sediments over the screened interval. Ideally, 3H/3He ages would represent a flow-weighted average over the entire thickness of the aquifer. Despite the uncertainty outlined above, 3H/3He age-dating is generally considered to be the most robust of the groundwater age-dating methods used in this study because the 3H atom is physically incorporated in the water molecule and is, therefore, a conservative tracer, unlike SF6 and CFCs, which are subject to contamination and degradation in some chemical environments. 4.3.3 Sulfur Hexafluoride Measured SF6 concentrations range from 0 to 11 fMol/kg (Table 8). Calculated mixing ratios range from 0 to 31 pptv. The current atmospheric mixing ratio of SF6 is 9.26 pptv (January 2017). Samples with no measurable SF6 presumably recharged before significant concentrations of SF6 were introduced to the atmosphere and are considered "premodern." Apparent ages range 31 from 0 to premodern, and correspond to recharge years from 2017 to pre-1953. Samples are considered "contaminated" if the calculated mixing ratio is greater than what is possible simply due to air-water equilibrium exchange during recharge (in other words, if the calculated mixing ratio is greater than the current atmospheric mixing ratio). Ten of the 30 samples are contaminated, up to three times the current (January 2017) atmospheric mixing ratio. Samples were initially analyzed for SF6 for comparison to 3H/3He ages; however, extensive contamination of lower Moab-Spanish Valley samples precludes this. Ten samples had calculated mixing ratios above the theoretical maximum for mere air-water equilibration at the water table prior to recharge. Additional samples, for which an SF6 apparent age was theoretically calculable, generally skew younger than the corresponding 3H/3He apparent ages (Figure 12), suggesting these are also affected by an additional source of SF6, whether natural or anthropogenic. Replicates were run for all but three samples, where either the bottle or cap was cracked. Measured replicate concentrations are plotted in Figure 13. Standard deviations are within 10% for the majority (18 out of 30) samples. The agreement between replicate samples suggests that SF6 contamination is not due to error in sampling or analytical methods. Busenberg and Plummer (2000) described total SF6 below, 𝑆𝑆𝑆𝑆6 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 = 𝑆𝑆𝑆𝑆6 𝑒𝑒𝑒𝑒 + 𝑆𝑆𝑆𝑆6 𝑒𝑒𝑒𝑒𝑒𝑒 + 𝑆𝑆𝑆𝑆6 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 + 𝑆𝑆𝑆𝑆6 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 − 𝑆𝑆𝑆𝑆6 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 (15) where 𝑆𝑆𝑆𝑆6 𝑒𝑒𝑒𝑒 is due to equilibrium air-water exchange, 𝑆𝑆𝑆𝑆6 𝑒𝑒𝑒𝑒𝑐𝑐 is addition from excess air during the dissolution and incorporation of air bubbles into the groundwater during water table rise, 𝑆𝑆𝑆𝑆6 𝑡𝑡𝑡𝑡𝑡𝑡𝑡𝑡 is addition of natural SF6, 𝑆𝑆𝐹𝐹6 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 is the addition of anthropogenic concentration, and 𝑆𝑆𝑆𝑆6 𝑙𝑙𝑙𝑙𝑙𝑙𝑠𝑠 is any loss due to degradation. If terrigenic SF6 and SF6 from contamination and SF6 loss are small, and excess air is determined independently, it is possible to date young groundwater 32 with SF6. Loss of SF6 was assumed to be negligible (and any loss of SF6 would make the contamination due to terrigenic or anthropogenic sources even more pronounced), so we focused on the potential sources of terrigenic or anthropogenic contamination. There is a strong spatial element to the contamination, relating to the lower valley and the wetland. Samples indicated in orange (Figure 12) were taken from wells within the wetland boundaries, except the U10 well, which is just upgradient (Figure 3). Gross contamination is not limited to the wetland; but, in general, contamination increases toward the Colorado River, especially when considered in relationship to samples taken for the USGS partner study in the upper valley (Figure 14). Many of the contaminated samples are from wells containing brine or brine-affected water; however, contamination does not appear to be inherently associated with the brines because the true brines (U4 and U5) contain essentially no SF6 (Figure 15). Since the deep brines at U4 and U5 do not exhibit any SF6, it seems unlikely that there is a natural terrigenic source, and more plausible that recent contamination is only affecting the shallower wells. Potential anthropogenic sources include the Ferrellgas underground injection site and the Atlas Mill Tailings. The Ferrellgas site was shut down in 2003 after failing mechanical integrity testing. The source of SF6 in the wetland remains an interesting research question, but is beyond the scope of this study. 4.3.4 Chlorofluorocarbons Measured CFC data are summarized in Table 9; values ranged from 0.015-17.4 pMol/kg for CFC-11, 0-69.7 pMol/kg for CFC-12, and 0-0.385 pMol/kg for CFC-113. Mixing ratios were calculated based on an assumed recharge elevation of 1500 m and recharge temperatures from dissolved noble gas analysis also based on a 1500 m recharge elevation. Mixing ratios ranged from 1.1-990 pptv for CFC-11, 0-20.2 pptv for CFC-12, and 0-72.1 33 pptv for CFC-113 (Table 10). These mixing ratios resulted in apparent ages that ranged from premodern to 0 years for CFC-11 and CFC-12, and 32 to 74 for CFC-113 (Table 11). Except for two samples (U7 and U15) that fall along the piston flow mixing line, CFC-11 appears to be degraded when plotted against CFC-12 (Figure 16). CFCs, especially CFC-11, are known to degrade in some anaerobic environments (Plummer and Busenberg, 2000). CFC-12 and CFC-113 concentrations are somewhat consistent along a piston flow model for samples older than approximately 35 years (Figure 16); the younger samples may have been affected by loss of CFC-113 due to sorption (Plummer and Busenberg, 2000). CFC-12 is generally considered to be the most stable CFC, and the most useful for groundwater age-dating. 4.3.5 Stable Isotopes Stable isotope data are presented in Table 12. Isotopic compositions range from -15 to 13.5 for δ18O and -108.8 to -101.5 for δ2H. The isotopic distribution was plotted in reference to the global meteoric water line (GMWL) and the Utah meteoric water line (UMWL) in Figure 17. All of the samples appear to be of meteoric origin, based on correlation with the GMWL. Several samples (namely, the brines) plot below the GMWL, an indication of evaporative enrichment. GCGA water is defined by Gardner (2004) as having δ18O of -14.5 to -15.0. Many of the samples appear to have been sourced from precipitation occurring at high elevation, similar to the GCGA end member. 4.4 Mill Creek Seepage (Bromide Tracer Test) The injected bromide concentration in the stream reached steady-state after approximately 4 hours (Figure 18). During the 18-hour period of steady-state, the bromide concentration in the stream increased steadily over time; this is a result of the stream flow declining steadily over the same time period, as the test was conducted on the tail-end of a rain event, as indicated by transducer data (Figure 19). 34 A synoptic sampling campaign (Figure 20) showed a gain of approximately 2.7 cfs below the confluence of Mill Creek and Pack Creek (due to inflow of Pack Creek), but less than 0.1 cfs of gain between Pack Creek and the Colorado River. The streamflow measurements that prompted the bromide tracer test indicated about 1 cfs of gain into a stream with a total flow rate of approximately 10 cfs. The measured gain is approximately 10 % of the total streamflow, which is near the error associated with manual streamflow measurements (approximately 5 %). It is, however, possible that the measured gain was real, but that it occurs intermittently. 35 7000 Cooper-Jacob Method from Specific Capacity 6000 Theis Recovery Method Transmissivity, ft2/day 5000 4000 3000 2000 U28, U29 U25 U24 U23 U22 U21 U20 U19 U18 0 U26, U27 1000 Figure 7. Aquifer testing transmissivity results from wetland preserve monitoring wells (monitoring well locations shown in Figure 3), in square feet per day (ft2/day) using different analytical solutions; hatched fill indicates estimate from observation well. The average transmissivity is approximately 1000 ft2/day. 36 Figure 8. Potentiometric surface (water table) map showing groundwater flow generally to the northwest through the valley bottom toward the Colorado River. The potentiometric surface map was generated using water levels measured from alluvial wells; the contour interval is 5 feet. Given that groundwater flow is perpendicular to potentiometric surface contours, the map indicates that a large proportion of groundwater discharge to the Colorado River is occurring at the south end of the wetlands preserve, near Mill Creek. 37 0 5 10 Depth, feet below MP 15 U26 U28 20 25 30 35 40 45 - 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000 SpC, uS/cm Figure 9. Graph displaying specific conductivity (SpC) profiles at U26 (blue) and U28 (orange); SpC values are shown in microSiemens per centimeter (µS/cm). The profiles indicate a transition from fresh groundwater to brine around 30 feet below the measuring point (i.e., top of the well casing). 38 Figure 10. Map of groundwater hydrochemical type, based on major-ion chemistry. The three observed groundwater types include sodium chloride (Na-Cl, indicated in red), calcium sulfate (Ca-SO4, green), and calcium bicarbonate (Ca-HCO3, blue). Darker hues indicate higher total dissolved solids (TDS); brine is defined here as having greater than 100,000 mg/L TDS, brackish is less than 100,000 mg/L and greater than 1,000 mg/L, and fresh is less than 1,000 mg/L. 39 Figure 11. Map showing the tritium/helium-3 apparent ages from valley-fill aquifer samples (increasingly older samples are indicated by darker shades of blue), potentiometric contours (black lines) and groundwater flow direction (red arrow), and the sample groupings used to calculate a discharge rate to the Colorado River. The discharge calculation was performed by measuring the distance between the right and left sample groupings, and then dividing the distance by the age difference between the average age of each respective sample grouping. 40 40 35 30 U20 U22 SF6 Age 25 U32 20 U16l U10 15 U15 U14 10 U13 U31 U19 5 U21 U18 U23 0 0 10 20 3H/3He 30 40 50 60 Age (Closed System Equilibrium Model) Figure 12. Plot showing the relationship between tritium/helium-3 and sulfur hexafluoride (SF6) apparent ages. Samples collected more distant from the wetlands preserve (shown in blue) have calculated apparent ages that correlate closer to the tritium/helium-3 age 1:1 line compared to samples collected in or near the wetlands preserve (shown in orange). Many of the SF6 samples collected were "contaminated," predominately those collected in the wetland preserve, as indicated by calculated apparent ages that were impossibly young (i.e., negative age). The results of this plot indicate that SF6 contamination is spatially correlated with the wetlands preserve. 41 12 10 Run 1, fMol/kg 8 6 4 2 0 0 2 4 6 8 10 12 Run 2, fMol/kg Figure 13. Graph showing replicate sulfur hexafluoride (SF6) measured concentrations, shown in femtoMoles per kilogram (fMol/kg). In general, replicate concentrations are correlative, indicating that SF6 contamination is not a result of measurement error. Figure 14. Study area map showing measured sulfur hexafluoride (SF6) concentration in femtoMoles per kilogram (fMol/kg). Samples collected in and around the wetland preserve indicate SF6 contamination, meaning that the SF6 concentrations are above what is possible solely through atmospheric equilibration. The current atmospheric concentration of SF6 is approximately 2 fMol/kg. 42 43 45 40 35 SF6, pptv 30 25 20 15 10 5 U4 U5 0 100 1000 10000 100000 1000000 Total Dissolved Solids (TDS), mg/L Figure 15. Plot comparing the calculated sulfur hexafluoride (SF6) partial pressure, in parts per trillion by volume (pptv), to total dissolved solids (TDS), in milligrams per liter (mg/L). The samples from U4 and U5, both located in the wetland preserve, have TDS concentrations exceeding 100,000 mg/L (i.e., brine), and essentially no SF6. This finding suggests that the observed SF6 contamination is not inherently associated with brine, but rather to an unknown anthropogenic or terrigenic contamination source. 44 Figure 16. Tracer-tracer plots show concentrations of two tracers in relation to the PFM (piston flow model, blue line) and EMM (exponential mixing model, red line). The top left diagram shows that, generally, CFC-11 is degraded relative to CFC-12. Top right shows that CFC-113 is also degraded relative to CFC-12, especially for younger samples. The bottom panel shows that older samples fall along the PFM or EMM mixing lines, and that the other samples may be captured by a binary mixing model (a tie-line between the PFM model and EMM model). 45 Figure 17. Stable isotope results. Upper Panel: Study area map showing oxygen and hydrogen stable isotope sampling locations. The symbols represent the type of water (e.g., groundwater versus surface water) collected at each location: Glen Canyon Group Aquifer, blue circle; valleyfill aquifer, green circle; brine, red circle; Mill Creek, blue triangle; Pack Creek, yellow triangle; and Mill Creek below the confluence with Pack Creek, green triangle. Lower Panel: Plot of stable isotope concentration (per mil) for hydrogen and oxygen; the solid line represents the global meteoric water line, and the dashed line is the Utah meteoric water line. 11/20/2015 0:00 11/19/2015 12:00 11/19/2015 0:00 11/18/2015 12:00 11/18/2015 0:00 11/17/2015 12:00 11/17/2015 0:00 Bromide, mg/L 6 5 4 3 2 T1 T2 T3 1 T4 0 -1 Figure 18. Establishing steady state at transport sites 46 Figure 19. Transducer data 20-Nov 00:00 19-Nov 12:00 19-Nov 00:00 18-Nov 12:00 18-Nov 00:00 17-Nov 12:00 17-Nov 00:00 16-Nov 12:00 Transport 4, change in water level, inches Transport 3, change in water level, inches Transport 2, change in water level, inches Transport 1, change in water level, inches 47 2 1 0 -1 -2 2 1 0 -1 -2 2 1 0 -1 -2 7 6 5 4 3 2 1 0 -1 -2 Synoptic 5 4 Bromide, mg/L 3 Pre-Synoptic 2 Synoptic 1 0 0 -1 5 10 15 20 25 30 35 40 Sample order, from upstream to downstream Figure 20. Synoptic 48 Table 1. Transmissivity results, square feet per day Site ID U18 U19 U20 U21 U22 U23 U24 U25 U26 U27 † U28 U29 † † Cooper-Jacob method Specific Capacity Theis recovery method Average Standard Deviation 60 920 3700 1700 1200 - - 270 30 1500 4100 4100 120 1300 3200 1300 290 80, 150 110 630 310 - 6200 - - 1500 2000 2200 460 340 60 1900 640 500 5900 4100 90 1300 3000 1700 640 190 90 930 330 1000 5400 4100 30 250 690 350 390 100 30 690 250 520 920 0 Observation well 49 Table 2. Field parameters and alkalinity Site ID Temperature ºC U1 U2 U3 U4 U5 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U26 U27 U28 U29 U30 U31 U32 pH 6.6 6.2 7.2 6.5 6.2 7.0 6.8 6.8 7.0 7.3 7.0 6.9 6.9 7.1 6.9 6.7 6.7 6.7 7.2 7.1 7.0 7.1 6.6 8.1 7.0 7.4 6.8 6.9 7.0 6.8 6.6 6.9 Total Dissolved Gases mmHg 740 834 750 880 1,274 692 680 660 800 679 720 692 685 661 682 645 694 712 700 744 691 664 712 748 760 773 676 741 727 715 682 675 Dissolved Oxygen mg/L 0.2 0.3 0.3 0.4 0.1 1.8 3.5 1.0 1.8 0.9 1.5 - - - - - - 2.8 0.3 - 7.7 - 1.1 0.2 2.9 2.0 4.5 0.4 1.2 1.5 0.1 5.9 Dissolved Oxygen % 3 6 4 7 2 22 45 13 21 9 18 - - - - - - 32 4 - 86 - 13 3 33 22 48 6 12 18 2 72 Alkalinity mg/L as CaCO3 255 186 300 260 165 268 270 194 124 211 300 234 216 228 216 230 314 190 180 170 273 298 174 244 310 - 183 - 216 136 275 124 50 ODL, over detection limit 16.6 19.2 16.2 13.2 13.7 15.2 16.8 17.8 16.1 15.2 15.7 15.9 15.9 16.9 17.5 17.3 16.6 17.1 15.4 15.9 14.3 12.6 16.9 19.3 13.6 12.6 12.7 12.5 13.4 16.0 16.3 17.6 Specific Conductance uS/cm 30,535 90,105 11,930 ODL ODL 5,899 1,092 1,824 2,115 952 680 1,586 1,574 796 1,519 1,158 998 2,423 987 905 921 899 3,581 2,437 3,306 12,219 1,223 46,341 5,188 1,521 906 1,323 51 Table 3. Specific conductivity profiles at select sites Depth BMP feet 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 BMP, below measuring point µS/cm, microSiemens per centimeter Specific Conductivity µS/cm MW-09 3,482 3,515 3,513 3,515 3,517 3,518 3,520 3,520 3,520 3,522 3,520 3,517 3,517 3,514 3,518 3,665 3,878 4,174 6,422 7,652 9,083 12,028 12,873 13,130 13,295 13,391 13,454 13,589 13,631 12,219 11,272 MW-11 - - 8,466 8,567 8,591 8,568 8,572 8,582 8,584 8,590 8,591 9,073 9,537 10,147 10,246 10,877 11,347 12,419 13,970 18,125 20,251 22,550 27,131 34,912 36,966 41,050 42,414 46,341 - - - Table 4. Major ion results Site Name U1 U2 U3 U4 U5 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U27 U29 U30 U31 U32 Sodium Potassium Calcium Magnesium Chloride Sulfate Bicarbonate TDS mg/L 6,648 24,309 2,553 32,426 41,072 905 36 27 30 23 13 79 59 32 79 53 29 179 24 17 19 20 234 506 544 49 1,104 37 27 75 mg/L 67 179 52 667 969 12 3 4 4 2 2 4 4 3 6 5 3 6 2 2 2 2 7 4 9 5 13 3 2 6 mg/L 1,118 2,615 202 1,909 1,977 237 173 325 382 142 71 181 204 86 175 114 162 274 127 117 124 116 601 23 143 136 195 260 125 139 mg/L 206 693 68 554 664 128 42 82 106 45 41 82 78 39 72 64 28 111 43 42 42 42 114 5 74 55 68 52 33 66 mg/L 19,932 66,287 6,258 84,907 108,183 2,216 28 32 54 24 5 48 48 24 42 58 32 170 22 22 21 19 283 696 925 36 2,264 56 17 78 mg/L 683 4,273 882 5,352 6,136 951 292 1,131 1,405 302 36 704 754 181 701 326 167 1,239 281 209 143 112 2,561 219 440 345 605 816 152 538 mg/L 311 227 366 317 201 327 329 237 151 257 366 285 263 278 263 280 383 232 219 206 333 363 206 298 378 223 263 166 335 151 28,964 98,581 10,381 126,131 159,201 4,776 902 1,838 2,131 795 533 1,382 1,411 643 1,338 900 802 2,212 719 613 684 674 4,005 1,752 2,513 850 4,511 1,391 691 1,053 mg/L Charge Balance % -23 -22 -22 -23 -24 -17 6 -8 -7 3 4 -3 -5 0 -5 1 5 -6 4 9 8 7 -13 -10 -5 7 -12 -6 5 0 52 Table 5. Measured noble gas concentrations Site ID U1 U2 U3 U4 U5 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U27 U29 U30 U31 U32 † Neon (ccSTP/g) 1.9E-07 1.7E-07 2.1E-07 1.7E-07 1.7E-07 1.9E-07 1.9E-07 1.7E-07 1.7E-07 2.0E-07 2.3E-07 1.9E-07 1.9E-07 1.8E-07 1.8E-07 1.7E-07 1.7E-07 2.1E-07 2.1E-07 2.0E-07 2.0E-07 8.4E-07 2.1E-07 2.3E-07 2.0E-07 2.0E-07 2.0E-07 2.1E-07 2.1E-07 1.7E-07 +(ccSTP/g) 4E-09 3E-09 4E-09 3E-09 3E-09 4E-09 4E-09 3E-09 3E-09 4E-09 5E-09 4E-09 4E-09 4E-09 4E-09 3E-09 3E-09 4E-09 4E-09 4E-09 4E-09 2E-08 4E-09 5E-09 4E-09 4E-09 4E-09 4E-09 4E-09 3E-09 Argon (ccSTP/g) 3.2E-04 3.0E-04 3.7E-04 3.2E-04 3.3E-04 3.3E-04 3.3E-04 2.7E-04 3.0E-04 3.7E-04 3.9E-04 3.6E-04 3.6E-04 3.3E-04 3.2E-04 3.2E-04 3.1E-04 3.6E-04 3.5E-04 3.7E-04 3.4E-04 9.0E-04 3.6E-04 3.9E-04 3.6E-04 3.6E-04 3.8E-04 3.7E-04 3.6E-04 3.0E-04 +(ccSTP/g) 9E-06 9E-06 1E-05 1E-05 1E-05 1E-05 1E-05 8E-06 9E-06 1E-05 1E-05 1E-05 1E-05 1E-05 1E-05 1E-05 9E-06 1E-05 1E-05 1E-05 1E-05 3E-05 1E-05 1E-05 1E-05 1E-05 1E-05 1E-05 1E-05 9E-06 Krypton (ccSTP/g) 7.3E-08 7.1E-08 8.8E-08 7.8E-08 7.5E-08 7.6E-08 7.7E-08 6.0E-08 7.3E-08 8.2E-08 8.5E-08 8.0E-08 7.8E-08 7.4E-08 7.8E-08 7.3E-08 7.4E-08 7.1E-08 7.3E-08 7.6E-08 7.8E-08 1.4E-07 7.6E-08 8.3E-08 7.5E-08 7.7E-08 7.8E-08 7.9E-08 8.0E-08 7.2E-08 +(ccSTP/g) 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 3E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 7E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 4E-09 Xenon (ccSTP/g) 1.0E-08 8.8E-09 1.2E-08 9.9E-09 1.0E-08 1.0E-08 1.0E-08 8.5E-09 1.0E-08 1.2E-08 1.2E-08 1.1E-08 1.1E-08 1.1E-08 1.0E-08 1.0E-08 1.0E-08 1.0E-08 1.0E-08 1.1E-08 1.0E-08 1.5E-08 1.0E-08 1.0E-08 9.5E-09 1.0E-08 1.1E-08 1.1E-08 1.0E-08 9.0E-09 +(ccSTP/g) 5E-10 4E-10 6E-10 5E-10 5E-10 5E-10 5E-10 4E-10 5E-10 6E-10 6E-10 6E-10 6E-10 6E-10 5E-10 5E-10 5E-10 5E-10 5E-10 5E-10 5E-10 7E-10 5E-10 5E-10 5E-10 5E-10 6E-10 5E-10 5E-10 4E-10 Helium-4 (ccSTP/g) 3.1E-07 8.6E-06 7.0E-07 7.4E-06 9.4E-06 7.5E-08 4.7E-08 6.3E-08 5.0E-08 6.8E-08 1.1E-07 5.0E-08 5.4E-08 4.3E-08 4.5E-08 5.0E-08 4.2E-08 6.6E-08 7.2E-08 5.5E-08 4.8E-08 3.9E-07 8.7E-08 1.4E-07 1.2E-07 1.1E-07 3.6E-07 2.9E-07 5.0E-08 4.8E-08 +(ccSTP/g) R/Ra † 0.209 0.088 0.149 0.071 0.067 0.529 1.014 1.123 1.677 1.168 0.490 1.013 0.948 1.098 0.932 1.040 0.995 0.826 1.135 2.053 1.792 1.096 0.650 0.468 0.439 0.641 0.333 0.322 1.248 0.941 R/Ra, where R is the ratio of 3He/4He measured in the sample, and Ra is the ratio of 3He/4He in the atmosphere (1.384E-6) 53 54 Table 6. Closed-system equilibration (CE) model (Aeschbach-Hertig et al., 2000) results, assuming recharge elevation of 1500 m Site ID U1 U2 U3 U4 U5 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U27 U29 U30 U31 U32 Recharge Temperature ºC 12.1 16.3 8.7 13.0 11.6 12.9 13.6 19.0 12.4 10.0 11.3 11.1 11.8 9.8 12.2 13.0 12.5 15.7 16.6 13.2 16.3 30.0 15.7 18.6 17.3 15.1 12.1 14.8 14.9 15.1 Excess Air Fractionation Factor Sum of Chi Squared 0.001 0.119 0.029 0.130 0.134 0.048 0.132 0.001 0.001 0.070 0.062 0.143 0.141 0.000 0.051 0.131 0.017 0.132 0.125 0.138 0.128 0.153 0.133 0.130 0.127 0.133 0.143 0.137 0.134 0.116 0.00 0.94 0.74 0.95 0.96 0.84 0.86 0.00 0.00 0.81 0.70 0.85 0.84 0.00 0.92 0.93 0.93 0.75 0.73 0.79 0.78 0.13 0.75 0.64 0.75 0.77 0.79 0.73 0.76 0.80 2E-02 2E+00 5E-08 8E-01 6E-01 8E-09 1E-02 6E-01 5E-01 2E-08 1E-08 7E-01 7E-01 3E-02 2E-07 4E-01 8E-08 6E-02 2E-07 8E-01 1E-02 2E-01 2E-01 2E+00 3E+00 4E-01 2E+00 4E-01 9E-01 3E-01 Table 7. Measured and calculated tritium results Site ID U1 U2 U3 U4 U5 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U27 U29 U30 U31 U32 Tritium +- TU TU 0.49 0.29 2.23 0.27 0.01 0.72 3.12 2.35 3.01 2.19 3.14 1.59 2.33 2.31 0.51 2.20 2.84 1.34 1.73 3.36 3.82 4.79 0.93 0.60 0.57 1.07 3.35 1.55 3.05 0.79 0.03 0.03 0.09 0.03 0.09 0.04 0.13 0.09 0.14 0.10 0.11 0.08 0.11 0.10 0.05 0.10 0.10 0.11 0.13 0.24 0.35 0.34 0.09 0.11 0.08 0.12 0.24 0.13 0.21 0.09 R/Ra Tritiogenic Helium-3 Terrigenic Helium-4 TU 0.209 0.088 0.149 0.071 0.067 0.529 1.014 1.123 1.677 1.168 0.490 1.013 0.948 1.098 0.932 1.040 0.995 0.826 1.135 2.053 1.792 1.096 0.650 0.468 0.439 0.641 0.333 0.322 1.248 0.941 4.80 163.32 14.60 65.84 68.60 -3.35 0.14 15.65 24.14 17.41 -0.52 4.31 0.91 1.66 0.31 3.70 0.80 2.89 18.69 36.36 19.63 110.80 4.78 2.42 3.18 13.18 33.15 19.25 9.40 2.38 Measured values: tritium, R/Ra; calculated values: 3Hetrit, 4Heterr, ΔNe, age † Computed negative ages are taken to be zero ∆Neon Age +- ccSTP/g % years years 2.69E-07 8.56E-06 6.57E-07 7.31E-06 9.37E-06 3.01E-08 -1.12E-09 2.08E-08 9.64E-09 2.02E-08 5.86E-08 6.67E-09 3.93E-09 -2.06E-09 2.97E-09 4.14E-09 9.37E-10 1.68E-08 2.44E-08 6.87E-09 -3.47E-09 1.70E-07 4.02E-08 7.98E-08 7.89E-08 6.62E-08 3.15E-07 2.39E-07 3.97E-09 7E-09 13 5 13 4 -9 13 25 16 6 17 30 10 30 10 6 12 -8 30 21 19 36 347 20 46 19 16 15 14 18 9 43 114 36 98 164 0† 1 37 39 39 0† 24 6 10 9 18 4 21 44 44 33 57 33 29 34 46 43 47 25 25 55 118 46 120 154 20 4 3 2 3 22 40 4 21 6 4 26 6 4 3 3 4 0 26 27 8 16 23 3 8 55 56 Table 8. SF6 results Well ID U1 U2 U3 U4 U5 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U27 U29 U30 U31 U32 SF6 Apparent Age (years as of 2017) C C C 42.1 45.6 C 1.6 C 16.1 15.1 15.6 0 5.6 11.1 12.1 18.6 C C 4.8 25.8 1.8 30.3 3.3 C Premodern 0 C C 6.3 18.3 Apparent Recharge Year C C C 1973.5 1970.0 C 2014.0 C 1999.5 2000.5 2000.0 2017.0 2010.0 2004.5 2003.5 1997.0 C C 2011.5 1990.5 2014.5 1986.0 2013.0 C Pre-1953 2017.0 C C 2010.0 1998.0 Calculated mixing ratio (pptv) 25.63 30.69 10.73 0.35 0.23 17.86 8.38 26.19 4.55 4.69 4.57 9.42 7.15 5.59 5.43 3.91 9.65 20.96 7.57 2.54 8.64 1.68 8.15 12.86 -1.04 9.36 13.69 20.02 7.12 4.17 Measured concentration (fMol/kg) 10.29 9.12 5.23 0.12 0.08 6.97 3.18 8.53 4.47 2.10 2.24 3.71 2.78 2.24 2.02 1.32 3.45 8.45 3.18 1.00 3.36 3.16 3.19 5.51 -0.39 3.59 5.62 8.21 2.78 1.28 "C" refers to "contaminated" samples, which have higher SF6 than possible from merely equilibrating with the atmosphere prior to recharge Table 9. Measured CFC results, pMol/kg Site ID U1 U2 U3 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U27 U29 U30 U31 U32 CFC-11 pMol/kg 0.051 0.089 0.066 0.503 1.904 4.269 0.029 0.056 0.172 12.719 0.878 17.373 3.863 9.715 0.194 4.501 0.769 0.486 0.381 0.036 1.348 1.314 0.892 0.015 1.891 0.449 0.134 2.002 +pMol/kg 0.005 0.099 0.021 0.009 0.024 0.211 0.003 0.042 0.130 0.577 0.011 1.195 0.025 0.104 0.001 0.055 0.714 0.475 0.565 0.025 0.027 0.741 0.826 0.002 0.067 0.016 0.004 0.028 CFC-12 pMol/kg 0.151 0.244 0.090 0.810 0.924 1.679 0.451 0.336 0.583 3.866 0.951 2.199 1.780 1.615 0.352 50.273 1.014 0.743 1.017 0.422 69.709 2.737 0.984 0.000 1.993 0.714 0.822 2.001 +pMol/kg 0.005 0.286 0.004 0.015 0.098 0.029 0.002 0.007 0.090 0.233 0.010 0.162 0.005 0.020 0.003 1.238 0.860 0.744 1.642 0.422 1.294 0.971 0.003 0.000 0.028 0.109 0.050 0.120 CFC-113 pMol/kg 0.003 0.018 0.003 0.003 0.113 0.045 0.000 0.000 0.385 0.180 0.024 0.165 0.152 0.142 0.011 0.124 0.055 0.042 0.140 0.021 0.035 0.150 0.079 0.006 0.155 0.030 0.008 0.163 +pMol/kg 0.004 0.025 0.004 0.005 0.002 0.002 0.000 0.000 0.647 0.006 0.001 0.011 0.001 0.002 0.008 0.005 0.055 0.042 0.229 0.014 0.001 0.035 0.047 0.004 0.010 0.007 0.011 0.025 57 Table 10. Calculated CFC mixing ratios Site ID U1 U2 U3 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U27 U29 U30 U31 U32 CFC-11 pptv 3.3 1.5 3.6 34.0 133.2 390.1 1.9 3.2 10.6 776.6 55.7 989.7 250.3 657.0 12.8 350.2 62.7 33.2 4.4 1.2 104.7 118.0 75.2 1.1 122.3 33.5 10.0 150.8 +pptv 0.3 0.1 1.1 0.6 1.7 19.3 0.2 2.4 8.0 35.3 0.7 68.1 1.6 7.0 0.1 4.2 58.2 32.5 0.01 0.8 2.1 66.5 69.6 0.2 4.3 1.2 0.3 2.1 CFC-12 pptv 37.1 12.4 18.8 207.2 244.0 563.1 112.4 74.6 138.0 903.8 230.1 483.2 438.8 413.8 88.2 14600.5 306.9 192.3 20.6 54.8 20199.3 903.9 307.1 0.0 490.7 199.6 230.6 564.8 +pptv 1.3 0.2 0.9 3.8 25.8 9.8 0.6 1.6 21.3 54.4 2.5 35.5 1.2 5.0 0.8 359.5 260.2 192.5 14.9 54.8 374.9 320.5 0.8 0 6.8 30.4 13.9 33.9 CFC-113 pptv 0.6 0.0 0.5 0.8 26.0 13.9 0.0 0.0 2.3 35.7 5.0 30.4 32.3 31.7 2.3 32.2 15.0 9.5 2.3 2.0 9.2 72.1 22.3 1.4 32.9 7.4 1.9 40.8 +pptv 0.8 0 0.7 1.1 0.4 0.7 0 0 3.3 1.1 0.1 2.0 0.3 0.5 1.6 1.2 15.0 9.5 0.3 1.3 0.3 46.7 13.3 1.0 2.1 1.8 2.7 6.3 58 Table 11. CFC apparent age results Site ID U1 U2 U3 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U27 U29 U30 U31 U32 † CFC-11 1955 1952 1955 1966 1976 2014 1953 1954 1959 2014 1970 2014 1988 2014 1961 2014 1965 1961 1956 Pre-1940 1974 1974 1967 1952 1975 1966 1960 1978 Recharge Year CFC-12 1960 1953 1956 1974 1976 2014 1969 1966 1970 2014 1975 1991 1987 1986 1967 2014 1988 Pre-1940 Pre-1940 Pre-1940 2014 2014 1980 Pre-1940 1990 1974 1975 2008 CFC-113 1949 1943 1949 1950 1981 1976 1943 1943 1952 1983 1969 1982 1983 1982 1958 1983 1963 1961 1963 1959 1973 1985 1979 1956 1983 1971 1952 1984 CFC-11 63 65 62 51 41 4 64 63 59 4 48 4 29 4 56 4 52 56 62 Premodern 43 43 50 66 42 51 57 39 +0.41 0.25 1.18 0.24 0.00 0.00 0.24 2.22 3.64 0.00 0.00 0.00 0.24 0.00 0.00 0.00 9.75 9.15 0.00 - 0.25 8.41 10.08 0.41 0.25 0.25 0.24 0.25 Apparent Age (years)† CFC-12 +57 0.24 64 0.25 62 0.41 43 0.24 41 1.41 4 0.00 48 0.24 51 0.22 47 1.24 4 0.00 42 0.24 27 2.55 30 0.00 31 0.47 50 0.00 4 0.00 29 25.75 Premodern - Premodern - Premodern - 4 0.00 4 0.00 37 0.00 Premodern - 27 0.75 43 1.48 42 0.85 9 9.53 CFC-113 68 74 69 67 36 41 74 74 65 34 48 35 35 35 59 35 55 56 54 58 44 32 39 61 35 46 65 33 +8.49 0.00 7.78 9.43 0.25 0.41 0.00 0.00 13.20 0.24 0.24 0.62 0.00 0.24 10.85 0.50 19.50 17.75 1.08 9.26 0.00 1.65 4.26 9.31 0.50 1.68 12.49 1.03 As of 2017 59 60 Table 12. Stable isotope results, permil Site ID U1 U2 U3 U4 U5 U6 U7 U8 U9 U10 U11 U12 U13 U14 U15 U16 U17 U18 U19 U20 U21 U22 U23 U24 U25 U27 U29 U30 U31 U32 δ18O ‰ -14.7 -14.0 -13.1 -13.5 -13.3 -14.3 -14.3 -14.8 -14.5 -14.8 -14.6 -14.4 -14.5 -14.6 -14.7 -14.5 -14.2 -14.4 -14.7 -14.6 -14.5 -14.5 -14.6 -14.8 -14.7 -14.7 -13.8 -14.7 -14.5 -15.0 δ 2H ‰ -109 -108 -101 -105 -105 -107 -106 -109 -108 -108 -107 -106 -107 -107 -108 -106 -105 -106 -108 -107 -106 -106 -107 -109 -108 -108 -103 -108 -106 -109 5 DISCUSSION 5.1 Conceptual Groundwater Model The valley-fill aquifer is an unconfined aquifer situated in Quaternary alluvial and colluvial sediments that have accumulated in the elongated basin of Spanish Valley. The collapsed salt anticline valley cuts off the thick Glen Canyon Group sandstones that rim the sides of the valley; the cliff-forming valley walls form physical boundaries of the aquifer on its northeastern and southwestern sides. At the distal end of the system, the shallow brines create a boundary to freshwater flow and essentially mark the bottom of the freshwater, valley-fill aquifer. Results of the electrical resistivity survey demonstrated, at least qualitatively, that the lens of fresh groundwater discharging to the Colorado River is thickest near Mill Creek, that brines are shallow in the north (i.e., the uppermost freshwater layer is thin or nonexistent), and that these are separated by a brackish transition zone in the middle (Figure 4). The potentiometric surface map shows that groundwater discharging near Mill Creek comes from Spanish Valley, whereas groundwater discharging to the north comes from the springs and groundwater from the ridge (Figure 8). The bromide tracer test indicated that essentially no groundwater (less than 0.1 cfs) is gained by Mill Creek in the wetland above the Colorado River. Valley-fill aquifer water must discharge either through evapotranspiration on the wetland or subsurface groundwater discharge to the Colorado River. Gardner (2004) characterized the groundwaters in the wetland into three categories according to geochemical observations. He defined GCGA water as having δ18O of -14.5 to -15.0 ‰, 3H of less than 2 TU, and R/Ra less than 1; the valley-fill as having δ18O of -14.0 to -15.0 ‰, H of 6.5 to 17.5 TU, and R/Ra greater than 1; and the brine as having δ18O of -13 to -13.5 ‰, 3 62 very low 3H, and R/Ra less than 0.1. Major ion chemistry reveals three geochemical water types: high-TDS sodium-chloride (brines), moderate-TDS calcium-sulfate, and low-TDS calcium-bicarbonate. Results indicate shallow brines in the wetland to the north and fresh groundwater near Mill Creek to the south, even at similar depths, corroborating the results of the electrical resistivity survey. Outside the wetland, samples from deeper wells exhibit the calcium-sulfate signature typical of Pack Creek surface water, suggesting that loss from Pack Creek is a large contributor of recharge to the valley-fill aquifer. Steiger and Susong (1997) reported that upper Moab-Spanish Valley where Pack Creek comes into the valley was a significant recharge zone for the valley-fill aquifer, and Pack Creek is known to be a losing stream that is dry in the upper valley much of the year. The low-TDS calcium-bicarbonate geochemical signature is observed in shallow wells on the north side of Pack Creek. GCGA waters are characterized by the low-TDS calciumbicarbonate signature (Mill Creek above the valley is fed by GCGA groundwater); however, these samples lack environmental tracer compositions expected from GCGA groundwater. Many of the samples had high elevation δ18O signatures; however, these waters do not have the "age" of GCGA water as given by 3H, R/Ra, and 4Heterr. Sumsion (1971) postulated that the valley-fill aquifer was recharged primarily by GCGA water from the northeast; however, we did not find significant amounts of "old" GCGA water in the valley-fill aquifer in lower Moab-Spanish Valley. The low-TDS calcium-bicarbonate samples could result from either loss from Mill Creek or runoff from GCGA springs along the northern valley wall whose dissolved noble gas signatures have had time to re-equilibrate. 5.2 Groundwater Discharge to Colorado River Throughout the western United States, groundwater budgets rely heavily on groundwater discharge estimates, because in many systems, discharge is spatially focused and readily 63 measured (e.g., from wells, springs, or small streams). In Spanish Valley, some groundwater may be discharging directly into a very large river, the Colorado, whose flow is not measurably changed by Spanish Valley discharge. Sumsion (1971) estimated that approximately 8,000 acre-feet of groundwater per year flowed to the Colorado River through the subsurface. Gardner (2004) used a Darcy Flux calculation along the length of the valley adjacent to the Colorado River to estimate between 100 and 1,500 acre-ft per year of groundwater discharge to the Colorado River. Two independent methods were implemented during this study to estimate the amount of groundwater entering the Colorado River from the subsurface, to follow up on the discrepancy between Sumsion (1971) and Gardner (2004). The first, the Darcy Flux method, uses transmissivity estimates from aquifer testing, the measured potentiometric surface, and the concept of flownet theory to estimate the amount of subsurface discharge to the Colorado using physical properties of the aquifer. The second method, the age gradient method, uses changes in apparent groundwater ages calculated from geochemical analyses of environmental tracers to calculate flow to the Colorado River, where the apparent age difference between samples gives us a direct measurement of the linear groundwater velocity along the flow path between sample sites. 5.2.1 Darcy Flux Method Using Darcy's law alone, the cross-sectional area of flow is required to calculate discharge. Because the thickness of the aquifer is not well defined, flownet theory is applied. A flownet is a two-dimensional graphical representation of groundwater flow, valid for steady-state conditions. A flownet is constructed of equipotential lines (hydraulic head contours), which are perpendicularly intersected by flowlines (or streamlines). The flownet in Figure 18 was generated from known hydraulic head contours (see section 4.1.2) as upper and lower boundaries. According to flownet theory, the total discharge through a flownet is calculated by 64 𝑄𝑄 = 𝑇𝑇∆𝐻𝐻𝑛𝑛𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 (16) where 𝑇𝑇 is transmissivity, ∆𝐻𝐻 is the hydraulic head contour interval, and 𝑛𝑛𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 is the total number of flowtubes in the flownet. Using a transmissivity of 1,000 ft2/day (section 4.1.1), ΔH of 5 ft in a flownet that generates 8 flowtubes, the total discharge to the system (𝑄𝑄) is 40,000 ft3/day, or approximately 300 acre-ft/year. 5.2.2 Age Gradient Method Assuming piston flow, the apparent age difference between samples along the same flow path gives a direct measure of the average horizontal linear velocity (v) between those sample sites, given by 𝑣𝑣 = 𝑑𝑑 ∆𝑇𝑇 (17) where 𝑑𝑑 is the distance along the flowpath between sample sites, and ∆𝑇𝑇 is the age difference between samples. Average linear velocity (𝑣𝑣) is related to specific discharge (𝑞𝑞) through porosity (𝑛𝑛), by 𝑞𝑞 = 𝑣𝑣𝑣𝑣 The age difference between samples (ΔT) was determined using two "clusters" of samples, approximately 2 miles apart (Figure 11). These samples were determined to be appropriately related because they lie along similar flowpaths and have similar geochemistry (Figure 10). Using 3H/3He apparent ages, the age difference between the upper and lower clusters was (18) 65 determined to be 30 ± 14 years. The age difference was determined to be 28 ± 16 years using CFC-12 analysis. SF6 apparent ages were discounted because of contamination (section 4.3.2), and CFC-11 and CFC-113 were discounted because of degradation and sorption, respectively (section 4.3.3). The distance was determined to be 9,000 ± 1,500 ft. The error quoted in the distance includes uncertainty introduced by spatial variations within a cluster of samples from which discharge was calculated. Assuming a porosity (n) of 0.3, the resulting specific discharge (q) is 90 ± 45 ft/year for 3 H/3He, and 96 ± 48 ft/year for CFC-12. Assuming an aquifer width of 5,000 ft and thickness of 100 ft, the resulting volumetric discharges according to 3H/3He and CFC-12 apparent age data are approximately 1,000 and 1,100 acre-ft/year, respectively. 66 Figure 21. Map depicting a flownet used to calculate the Darcy flux discharge to the Colorado River through the wetland preserve. A flownet consists of equipotential lines (dark blue) and perpendicular flowlines (light blue); flowtubes are the regions between flowlines (eight total flowtubes). 6 CONCLUSION Contrary to the conceptual model outlined by Sumsion (1971) that portrays the valley-fill aquifer as primarily recharged by GCGA water along the northern wall of the valley, we did not find "old" GCGA water in samples taken in the lower valley except for hints of mixtures that could contain modest amounts of GCGA water. Our data suggest that most of the GCGA water is discharging at the Moab City springs, and that the primary source of recharge to the valley-fill aquifer is surface water from Pack Creek and Mill Creek, which was precipitated at high elevation or at least above the study area. Two independent methods were used to estimate the quantity of groundwater discharge in the wetland, either directly to the Colorado River or via evapotranspiration by wetland phreatophytes. The Darcy flux method yielded an estimate of approximately 300 acre-ft per year, while the age-gradient method yielded approximately 1,000 acre-ft per year. These estimates are significantly less than the previous estimate by Sumsion (1971) of 8,000 acre-ft per year, but agree with the estimates made by Gardner (2004) of 100 to 1,500 acre-ft per year. The bromide tracer test indicates that less than 0.1 cfs of groundwater is escaping the flow system through Mill Creek above the Colorado River. Our data indicate that the volume of water requested for water-rights transfer by the SJSVSSD does not exist in the valley-fill aquifer. Better understanding of the valley-fill aquifer system could be gained by gauging the streams that are the main inputs to the aquifer, Mill and Pack Creeks, to better constrain gains and losses. APPENDIX A BROMIDE TRACER TEST 69 A bromide tracer test was performed along Mill Creek near the Colorado River to evaluate whether groundwater was discharging into Mill Creek before reaching the Colorado River. The need for the tracer test was prompted by flow measurements taken with a SonTek FlowTracker Handheld-ADV (Acoustic Doppler Velocimeter) on an approximately 1-mile reach of lower Mill Creek (Figure 22) that indicated a gain of approximately 1 cfs. A bromide tracer injection was designed to locate and quantify the gain. A bromide tracer injection uses a concentrated solution of sodium bromide (NaBr) injected at a constant, known rate, whereby any dilution in measured concentrations of samples taken downstream indicates the occurrence of groundwater inflow (seepage) into the stream. The bromide injectate was created by mixing six 55 lb bags of NaBr with approximately 60 gallons of water to create a solution with a concentration of approximately 500,000 mg/L. Stream measurements in the target area were approximately 9.8 cfs. The goal concentration for the samples was 3 mg/L. The pump rate was set to approximately 100 mL/min to achieve the target concentration (𝐶𝐶𝑠𝑠 ) determined by following equation: 𝐶𝐶𝑖𝑖 𝑋𝑋𝑖𝑖 = 𝐶𝐶𝑠𝑠 𝑋𝑋𝑠𝑠 (19) where 𝐶𝐶𝑖𝑖 is the bromide concentration of the injectate, 𝑋𝑋𝑖𝑖 is the rate of injection, 𝐶𝐶𝑠𝑠 is the bromide concentration in the stream, and 𝑋𝑋𝑠𝑠 is the maximum expected stream discharge. An injection site was selected approximately one quarter mile upstream from the target area (Figure 23) to insure the injectate was fully mixed into the stream water before reaching the target area. A funnel was built using stream stones to funnel water just below the injectate to facilitate mixing in the rocky, shallow stream. Three transport sites were established before the test to show when the bromide concentration in stream reached steady-state. At each site, an ISCO auto-sampler was set to collect a stream sample once every hour. Locations of the transport sites are shown in Figure 20. 70 A presynoptic was preformed to establish a baseline bromide concentration in the stream by identifying any natural sources of bromide; 12 presynoptic samples were collected. These values would be subtracted from synoptic values after analysis. The solution was injected into Mill Creek at a constant rate for a duration of 30 hours. The bromide concentration in the stream reached steady state after approximately 4 hours, determined by time series of samples collected at four locations (transport sites) along the stream (Figure 18). After steady-state conditions were achieved, a synoptic of samples was collected at 24 locations along the stream. After steady-state conditions were assumed to have been reached in the stream, a synoptic was taken where samples were collected from the same locations as in the presynoptic. The samples were analyzed for bromide at the geomicrobiology lab at the University of Utah, in Salt Lake City, Utah. Samples were analyzed on a Metrohm 883 Basic IC Plus ion chromatograph at the Geomicrobiology Laboratory at the University of Utah in Salt Lake City, Utah. A standard was run every 10 samples, which allowed the concentrations to be corrected for instrument drift (Figure 24). Figure 22. Flow measurements along Mill Creek that prompted the tracer test 71 Figure 23. Map of bromide tracer test; location of injection site, transport sites, pre-synoptic, and synoptic Check Standard Corrected vs Uncorrected 120 115 110 Bromide (mg/L) 105 Corrected 100 Uncorrected 95 90 85 80 72 Figure 24. Sample correction (instrumental drift) APPENDIX B AQUIFER TESTING 74 U18 Pump Test 2 5 1.5 10 1 15 0.5 20 25 Pumping Rate (gpm) Drawdown (feet) 0 0 0 30 60 90 120 150 180 210 Time (minutes) 0 1 2 3 4 5 6 7 8 6 5 4 3 2 1 0 0 30 60 90 Time (minutes) 120 150 180 Pumping Rate (gpm) Drawdown (feet) U19 Pump Test 75 U20 Pump Test 6 5 1 4 2 3 2 3 1 4 Pumping Rate (gpm) Drawdown (feet) 0 0 0 30 60 90 120 150 180 210 Time (minutes) 0 6 1 5 2 4 3 3 4 2 5 1 6 Pumping Rate (gpm) Drawdown (feet) U21 Pump Test 0 7 0 30 60 90 120 150 180 Time (minutes) U22 Pump Test 2 2 1.5 4 1 6 8 0.5 10 0 12 0 30 60 90 120 Time (minutes) 150 180 210 Pumping Rate (gpm) Drawdown (feet) 0 76 0 1.2 5 1 0.8 10 0.6 15 0.4 20 0.2 25 0 0 30 60 90 120 150 Pumping Rate (gpm) Drawdown (feet) U23 Pump Test 180 Time (minutes) -2 0 2 4 6 8 10 12 14 16 18 1.4 1.2 1 0.8 0.6 0.4 0.2 Pumping Rate (gpm) Drawdown (feet) U23 Pump Test Redo 0 0 30 60 90 120 150 180 210 Time (minutes) -1 1 3 5 7 9 11 13 15 1.2 1 0.8 0.6 0.4 0.2 0 0 30 60 90 120 150 Time (minutes) 180 210 240 270 Pumping Rate (gpm) Drawdown (feet) U24 Pump Test 77 U25 Pump Test 2.5 Drawdown (feet) 0 2 1 2 1.5 3 1 4 0.5 5 Pumping Rate (gpm) -1 0 6 0 30 60 90 120 150 180 Time (minutes) 0 1.0 1 0.8 2 0.6 3 0.4 4 0.2 5 Pumping Rate (gpm) Drawdown (ft) U26, U27 Pump Test 0.0 0 30 60 90 120 150 180 210 Time (minutes) 0.0 6 0.2 5 0.4 4 0.6 3 0.8 2 1.0 1 1.2 1.4 0 0 30 60 90 120 Time (minutes) 150 180 210 Pumping Rate (gpm) Drawdown (ft) U28, U29 Pump Test REFERENCES Aeschbach-Hertig, W., F. Peeters, U. Beyerle, and R. Kipfer, 1999, Interpretation of dissolved atmospheric noble gases in natural waters: Water Resources Research, v. 35, p. 27792792. Aeschbach-Hertig, W., F. Peeters, U. Beyerle, and R. Kipfer, 2000, Palaeotemperature reconstruction from noble gases in ground water taking into account equilibration with entrapped air: Nature, v. 405, p. 1040-1044. Brown, R., J. Ferris, C. Jacob, D. Knowles, R. Meyer, H. Skibitzke, and C. Theis, 1963, Methods of determining permeability, transmissibility and drawdown: US Geological Survey Water-Supply Paper, v. 1536, p. 341. Bullister, J. L., D. P. Wisegarver, and F. A. Menzia, 2002, The solubility of sulfur hexafluoride in water and seawater: Deep Sea Research Part I: Oceanographic Research Papers, v. 49, p. 175-187. Busenberg, E., and L. N. Plummer, 2000, Dating young groundwater with sulfur hexafluoride: Natural and anthropogenic sources of sulfur hexafluoride: Water Resources Research, v. 36, p. 3011-3030. Clarke, W. B., W. Jenkins, and Z. Top, 1976, Determination of tritium by mass spectrometric measurement of 3He: The international journal of applied radiation and isotopes, v. 27, p. 515-522. Cooper, H. H., and C. E. Jacob, 1946, A generalized graphical method for evaluating formation constants and summarizing well‐field history: Eos, Transactions American Geophysical Union, v. 27, p. 526-534. Craig, H., 1961, Isotopic variations in meteoric waters: Science, v. 133, p. 1702-1703. Doelling, H. H., 1983, Observations on Paradox Basin salt anticlines, in Averett, W.R., editor, Northern Paradox Basin-Umconpaghre uplift: Grand Junction Geological Society Field Trip, p. 81-90. Doelling, H. H., M. L. Ross, and W. E. Mulvey, 2002, Geologic map of the Moab 7.5'quadrangle, Grand County, Utah, Utah Geological Survey, Utah Department of Natural Resources. Driscoll, F. G., 1986, Groundwater and wells: St. Paul, Minnesota: Johnson Filtration Systems Inc., 1986, 2nd ed., v. 1. Fillmore, R., 2010, Geological Evolution of the Colorado Plateau of Eastern Utah and Western Colorado, Including the San Juan River, Natural Bridges, Canyonlands, Arches, and the Book Cliffs, University of Utah Press. 79 Gardner, P. M., 2004, Environmental Tracer Investigation of Groundwater Conditions at the Scott M. Matheson Wetlands Preserve Near Moab, Utah, University of Utah. Halford, K. J., W. D. Weight, and R. P. Schreiber, 2006, Interpretation of transmissivity estimates from single‐well pumping aquifer tests: Ground Water, v. 44, p. 467-471. Heilweil, V. M., D. S. Sweetkind, and S. J. Gerner, 2014, Innovative environmental tracer techniques for evaluating sources of spring discharge from a carbonate aquifer bisected by a river: Groundwater, v. 52, p. 71-83. Kaufman, S., and W. Libby, 1954, The natural distribution of tritium: Physical Review, v. 93, p. 1337. Koh, D.-C., L. N. Plummer, E. Busenberg, and Y. Kim, 2007, Evidence for terrigenic SF 6 in groundwater from basaltic aquifers, Jeju Island, Korea: Implications for groundwater dating: Journal of Hydrology, v. 339, p. 93-104. Lovelock, J. E., 1971, Atmospheric fluorine compounds as indicators of air movements: Nature, v. 230, p. 379. Maiss, M., and I. Levin, 1994, Global increase of SF6 observed in the atmosphere: Geophysical Research Letters, v. 21, p. 569-572. Manning, A. H., and D. K. Solomon, 2003, Using noble gases to investigate mountain-front recharge: Journal of Hydrology, v. 275, p. 194-207. Pataki, D. E., S. Bush, P. Gardner, D. K. Solomon, and J. R. Ehleringer, 2005, Ecohydrology in a Colorado River riparian forest: implications for the decline of Populus fremontii: Ecological Applications, v. 15, p. 1009-1018. Plummer, L. N., and E. Busenberg, 2000, Chlorofluorocarbons, Environmental tracers in subsurface hydrology, Springer, p. 441-478. Solomon, D. K., and P. G. Cook, 2000, 3h and 3he, Environmental tracers in subsurface hydrology, Springer, p. 397-424. Stute, M., and P. Schlosser, 2000, Atmospheric noble gases, Environmental tracers in subsurface hydrology, Springer, p. 349-377. Sumsion, C., 1971, Geology and water resources of the Spanish Valley area: Grand and San Juan Counties, Utah: Utah Department of Natural Resources Technical Publication, v. 32, p. 45. Theis, C. V., 1935, The relation between the lowering of the Piezometric surface and the rate and duration of discharge of a well using ground‐water storage: Eos, Transactions American Geophysical Union, v. 16, p. 519-524. Wilson, R. D., and D. M. Mackay, 1993, The use of sulphur hexafluoride as a conservative tracer in saturated sandy media: Ground Water, v. 31, p. 719-724. http://www.noblegaslab.utah.edu/pdfs/cu_tube_sampling.pdf Reproduced with permission of copyright owner. Further reproduction prohibited without permission.