BernauGeochem.pub 1 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2023 Utah Geological Association Publication 51 ABSTRACT Over the past century, the Bonneville Salt Flats, which lies on the western edge of the Great Salt Lake wa- tershed, has experienced changing environmental conditions and a unique history of land use, including re- source extraction and recreation. The perennial halite salt crust has decreased in thickness since at least 1960. An experimental restoration project to return mined solutes began in 1997, but it has not resulted in anticipated salt crust growth. Here, primary observations of the Bonneville Salt Flats surface and subsurface brine chemis- try and water levels collected from 2013 to 2023 are reported. Spatial and temporal patterns in chemistry, fo- cused on density and water stable isotopes, are evaluated and compared with observations across seven periods of research spanning from 1925 to 2023. Declining salinity in the areas to the east of extraction ditches and south of Interstate 80 were observed. Brine extracted for potash production decreased in salinity as extraction rates increased. Between the years 1964 and 1997, brine in the shallow aquifer located beneath and to the east of the crust experienced a decrease in salinity. However, following this period, the salinity stabilized and sub- sequently increased. Salinity recovery was concurrent with declines in brine extraction and the salt restoration project, with the largest decrease in brine extraction being concurrent with the largest recovery in salinity. The specific impact of the restoration project on the brine salinity increase remains unclear. To the west, the shal- low aquifer in the area between the Silver Island Mountains and the salt crust has increased in salinity. This increase is accompanied by a decline in groundwater levels, which enables the underground movement of so- lutes from east to west, following a salinity gradient away from the saline pan. Over the past 25 years, water levels in the alluvial-fan aquifer along the Silver Island Mountains have markedly declined, leading to the ex- traction of increasingly more saline and isotopically heavier basinal waters are intriguing landscapes for indus- trial use. This change is concurrent with the onset of the salt restoration project, which relies on alluvial-fan aquifer waters. This article’s compilation of changes in groundwater chemistry provides an important resource for stakeholders working to understand and manage this dynamic and ephemeral evaporite system. It also of- fers an example of decadal-scale change in a highly managed Great Salt Lake watershed saline system. Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats, Utah   Jeremiah A. Bernau1,2, Brenda B. Bowen1, Evan L. Kipnis1, and Jory C. Lerback1,3 1Department of Geology and Geophysics, University of Utah, Salt Lake City, Utah, jeremiahbernau@gmail.com 2Utah Geological Survey, Salt Lake City, Utah 3Department of Earth, Planetary and Space Sciences, University of California, Los Angeles, California INTRODUCTION Saline pans, shallow depressions encrusted by evaporites where waters accumulate, provide an intri- guing example where groundwater level and chemis- try, climate, and anthropogenic activities converge. Here, several decades of chemical measurements, with a focus on density, are used to examine how the Bonneville Salt Flats (BSF) groundwater system changed in response to: 1) brine extraction for potash production, 2) alluvial-fan groundwater extraction for industrial uses, and 3) 25 years of an experimental brine “laydown” program to restore the saline pan (Figures 1 to 3) (Kipnis and Bowen, 2018). The laydown program uses alluvial-fan aquifer groundwa- ter to dissolve the potash mine’s halite (NaCl) by- product and transport it to the saline pan in hopes of restoring saline pan thickness and extent. Multi- decadal analyses of satellite imagery and reoccurring measurements of salt crust thickness show long-term declines in crust thickness and extent (Bowen and others, 2017; Bowen and others, 2018; Radwin and Bowen, 2021). One-third of the crust consists of hal- ite (NaCl) and two-thirds of it is gypsum (CaSO4·2H2O). The multi-decadal nature of research on BSF and the uniquely involved mix of stakehold- ers including racing enthusiasts, recreational visitors, the potash industry, researchers, and governmental managers make this site well-suited for examining the evolution of brine chemistry. This landscape is dy- namic with seasonal to decadal-scale changes in flooding and saline pan volume (Figure 2E and F) (Bowen and others, 2017). Here, this examination of long-term changes in brine chemistry, with ground- water levels as a secondary dataset, provides context for the relative impact of extraction and restoration 10.31711/ugap.v51i.143 2 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats Figure 1. Site overview. A. Off-angle International Space Station image of Bonneville Salt Flats, Great Salt Lake Desert, and Great Salt Lake (https://earthobservatory.nasa.gov/images/91765/bonneville-salt- flats). B) Areas of investigation. Sample areas (divided as zones moving east from west) and depths of different sampling wells noted in figure explanation. Inset C) shows locations of brackish water (BW) alluvial-fan aquifer production wells (A-A’) and nested observational wells (OW) (B-B’) along tran- sects. Primary production wells outlined by box in the middle of A-A’. D) Schematic of investigated aq- uifer intervals (not to scale). Basemap imagery from Earthstar Geographics. 3 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 Figure 2. Features influencing brine density and surface variation at BSF. A). Aerial photograph of pot- ash mine looking north to BSF, letters denote relative locations of B to D. B) Alluvial-fan aquifer wells collect brackish water that is used in mine operations and to create salt laydown. C) Brine collection ditch (6 m deep) (Bingham, 1980) east of BSF (looking north). D) Laydown brine being introduced to BSF’s southwest corner, person on right for scale. E and F) Time-lapse photos from BSF weather sta- tion on E) May 28, 2018, when the surface was flooded, and F) July 15, 2018, when the surface was des- iccated. More field and aerial imagery of BSF is available with the Utah Geological Survey Data Ar- chive system at https://geodata.geology.utah.gov/pages/search.php?search=%21collection129324. 4 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats Figure 3. Human activities (mining and laydown) that impact brine and solute mass through time. Changes in brine mass balance (A and B) in millions of tonnes (Mt) of NaCl (not brine), the density of extracted brines (C and D), and alluvial-fan aquifer (AF) groundwater extraction and groundwater levels (E). A) Annual values NaCl mass extracted from BSF’s western ditches, added to the saline pan through the laydown, and net annual NaCl balance. B) Cumulative net movement of NaCl onto BSF from the laydown. C) Changes in extracted brine salinity over time. Note seasonality in density; summer values used for long-term trend. Numbered red arrows added to B) and C) highlight notable inflec- tion points in data. In 2006 (point 1) the salinity of extracted brine began a long-term decline. Net cumulative brine contributions were neutral between 2005 and 2010 (point 2), and increased afterward; despite this increase, extracted brines did not begin to recover in salinity until 2015 (point 3). After 2017 (point 4) extracted brine salinity remains relatively high, while extraction is low. D) Correlation between pumping and extracted brine salinity (each dot repre- sents the average of 8 months of pumping and brine density values) from Aug. 2018 to Dec. 2020 (these values are re- ported as hours of pumping, ~0.85 acre-ft/hour). E) Alluvial-fan annual groundwater extraction and groundwater ele- vation (meters above sea level) change over time (well located at 2.3 km in Figure 12) (103 acre-ft ≈ 1.23*106 m3). The vertical dashed red line shows when laydown began, and the gray area on (A) and (D) emphasizes the period with ele- vated laydown volumes and lower than anticipated (light blue line in lower E) groundwater levels. Alluvial-fan extrac- tion data from https://www.waterrights.utah.gov/cgi-bin/wuseview.exe?Modinfo=WRUseage&wrnum=16-25. 5 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 activities on changing saline pan extent and volume (Figure 3). This improved knowledge will help guide plans to optimize the sustained use, including both mining and racing, of this landscape. METHODS Brine chemistry was characterized and groundwa- ter levels were measured at BSF between 2015 and 2023. Measurements prior to 2015 were compiled. This study involved extensive density measurement quality control, analyzing trends in individual wells, and analyzing data aggregated by area and aquifer. Here, groundwater levels are reported to provide con- text for observed changes; however, they are not the primary focus of this work. Data Sources Over Time Groundwater level and density data were divided over seven study periods that vary in length, intervals between periods, reported data, areas of investigation, and researchers (Figures 4 and 5). Nolan (1927) in- vestigated the composition of Great Salt Lake Desert (GSLD) brines. In 1925, Nolan made shallow borings across the GSLD and reported sample total dissolved solids (TDS), major ions, and groundwater levels. Nolan made major-ion measurements in the field through titration (for chloride) and by measuring the volume of precipitate formed after adding chemicals to the solution (for potassium and sulfate). Several of these sites occur on the southern edge of BSF and the area between the Silver Island Mountains and the sa- line pan crust (Figure 4B). Between 1946 and 1949, 23 alluvial-fan produc- tion wells (named BW for brackish water) were drilled to depths ranging between 32 and 111 m (Bernau and others, 2023a) (Figure 1C). BW wells flowed freely (1.1 to 9.5 m3/second) when they were drilled, with reported potentiometric surfaces between 1.5 and 6 m above surface level. BW well water chemistry was reported in aggregate. Two researchers investigated BSF chemistry be- tween 1964 and 1972 (Figure 4C). In 1964 brine sam- ples from shallow GSLD borings were analyzed for major ions, lithium, and TDS (Lindenburg, 1974). Between 1965 and 1967, Turk (1973) installed shal- low (<9 m depth) wells across and adjoining BSF’s crust, and measured groundwater levels and brine chemistry. In 1972, two brine chemistry samples from the alluvial-fan aquifer wells were collected (reported in the Water Quality Portal; Read and others, 2017). Between 1975 and 1981 the U.S. Geological Sur- vey performed two BSF studies (Figure 4D). Ground- water levels and chemistry were examined between 1975 and 1978 (Lines, 1978, 1979). In 1981 the U.S. Geological Survey Conservation Division measured brine chemistry from borings and wells (reported in White, 2002). The next study period occurred leading up to the onset of the laydown in the autumn of 1997. Two groups collected measurements between 1991 and 1997 (Figure 4E). The U.S. Geological Survey meas- ured BSF between 1991 and 1993 (Mason and others, 1995; Mason and Kipp, 1998). The Bureau of Land Management made annual measurements from a sub- set of wells between 1994 and 1997 (White, 2002). Between 1998 and 2006 the Bureau of Land Man- agement monitored brine chemistry and water levels in several wells to evaluate the laydown’s impact (Figure 4F) (White, 2002). Some unpublished Bureau of Land Management measurements collected be- tween 2003 and 2006 are compiled here (White, field notes and files including laboratory results, 1998 to 2014). Between 2003 and 2018, Shaw Environmental, Inc. conducted biannual measurements of groundwa- ter level and chemistry from various locations. These measurements, amounting to over 900 measurement sets, were gathered on behalf of the potash mine as part of its mine reclamation plan (Shaw Environmen- tal, 2020). While most sites were within the potash mine, many samples were from the BSF study area. These include samples from the alluvial fan aquifer, east of the extraction ditch, and south of Interstate 80 (I-80) areas. The samples for the period spanning from 2007 to 2012 exclusively originate from this re- port. Between 2015 and 2022, researchers from the University of Utah collected brine chemistry and groundwater level measurements, and between 2022 and 2023 the Utah Geological Survey collected simi- lar measurements (Figure 4G) (Penrod, 2016; Bowen and others, 2018; Kipnis and Bowen, 2018; Lerback and others, 2019; Kipnis and others, 2020; Bernau and Bowen, 2021; Bernau and others, 2023a). The Bureau of Land Management made groundwater level and density measurements independently (White, field notes and files including laboratory results, 1998 to 2014) and in collaboration with the University of Utah in 2015. Measurements From May 2016 to May 2020, a precipitation sample collector with internal electrical heating was installed at the potash mine. Mine staff monitored and collected precipitation samples regularly, offering storm-event-level resolution for collected precipita- 6 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats Figure 4. Spatial distribution of shallow brine aquifer density measurements from the Bonneville Salt Flats and sur- rounding area. A) Compilation from density measurements between 1925 and 2023. B to G) Locations of shallow brine aquifer density measurements over time. Kriging done with Empirical Bayesian Kriging (Geostatistical Ana- lyst) in ArcGIS. 7 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 Figure 5. Brine density measurements over time (~2000 data points). Measure- ments are separated by area and depth of investigation. Note the difference in vertical scales. Light red arrows added to highlight trends by area. 8 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats tion. Stable water isotopes of hydrogen (δ2H) and ox- ygen (δ18O) were measured at the SPATIAL Lab at the University of Utah with a Picarro L2130-i cavity ring-down spectrometer. Water isotope concentra- tions were calibrated to laboratory standards, and val- ues were reported as per mil (‰) relative to the V- SMOW scale. Between 2016 and 2021, ground and surface wa- ter samples were collected in acid-washed and deion- ized water-rinsed bottles and returned to the Universi- ty of Utah for major ion and stable isotope analyses. A subset of samples was analyzed for major ion con- centrations by external laboratories using inductively coupled plasma mass spectrometry (ICP-MS/AES) and ion chromatography as reported in Kipnis and others (2020). Samples collected in 2017 and 2020 were analyzed for major ion concentrations using a portable X-ray fluorescence spectrometer (pXRF) and calibrated to the methods described in Kipnis and oth- ers (2020). Calcium concentrations and total dis- solved solids were excluded from pXRF results due to data calibration challenges. Additional samples were collected in 2017 and 2018 for chemical analyses in- cluding radiocarbon (C14) and tritium (3H/3He) dating as reported in Lerback and others (2019). Samples collected in 2021 were measured for major ions using ICP-MS and ion chromatography at the University of Utah Earth Core Facility. Samples collected in 2022 were analyzed for major ion concentrations by exter- nal laboratories (Chemtech-Ford, Utah Public Health Laboratory). Brine density values were measured in field and lab settings with a Mettler-Toledo Densito 30PX. Laboratory measurements of density at 20°C were made following the procedures of Bernau and others (2023a) at 20 °C. Before 2022, when analyzing hydrogen (δ2H) and oxygen (δ18O) stable isotopes from brine samples, solutes were removed through cryogenic vacuum extraction before stable isotope measurement. Starting in 2022, vacuum extraction was not used for brine samples analyzed for stable isotopes. Data Quality Control Due to the diversity in data vintages, care was tak- en in reviewing data quality. Because of this, changes in density were primarily focused on as density meas- urements are relatively robust over time and are less susceptible to methodological changes (Bernau and others, 2023a). Salinity is directly proportional to brine density and the two terms are used interchange- ably here. In addition to data quality, seasonal chang- es in brine salinity were considered. Seasonal varia- tions in brine density, as observed from monitoring well brines and brines extracted for potash produc- tion, indicate that samples collected during cooler and wetter winter to spring months are more likely to have depressed density measurements (Figure 6). Measurements with higher density from warmer, drier months (July to September, preferably August), are preferred for long-term evaluation. While the impact of temperature on density measurements was taken into account and corrected for whenever feasible (with a change of approximately 0.01 g/cm3 observed between temperatures of 10 and 30 °C), it should be noted that warmer brines can dissolve more halite, re- sulting in higher densities (Bernau and others, 2023a). To mitigate the influence of dilution caused by flood- ing, it is advisable to utilize the highest recorded den- sity at a site during a study period for long-term com- parisons. Brine chemistry reporting varied across studies. Some studies only reported field density, others re- ported major ion chemistry and periodically TDS, whereas others reported laboratory density measure- ments in addition to the measurements above. Field densities with reported temperature (if available) were corrected to the density at 20°C using Equation 11 in Bernau and others (2023a). Following the methods in Bernau and others (2023a), major ion data were used to model density using the SpecE8 module of Geo- chemists’ Workbench® with the PHRQPITZ thermo- dynamic dataset (Pitzer, 1973; Harvie and others, 1980; Plummer and others, 1988; Bethke, 2013). Fi- nally, available measurements were utilized to estab- lish the correlation between TDS and density for BSF brines. Using the measured TDS data, a salinity for these brines was estimated. Chemical model-based estimates of density at BSF tend to underestimate density, indicating that major ion concentrations are typically underreported (Bernau and others, 2023a). When all of these measurement types and estimates of salinity were available, they were contrasted to delin- eate measurement quality. For long-term comparative analyses, laboratory density measurements were pri- oritized, then field density measurements, followed by chemically modeled density, and finally density estimated from TDS. An additional step in data quality control was made using site-based knowledge to assess data quali- ty and identify erroneous data to remove. For exam- ple, anomalously high density values (>1.22 g/cm3) are not possible at BSF given its brine composition and suggest measurement errors, such as suspended sediment increasing field density measurements. An additional consideration was unusually low density measurements. For example, some samples from the years 1991 to 1993 have unusually low reported den- sities. Samples that contain higher levels of sodium M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 9 Figure 6. Seasonality in density measurements across different sample areas. Data from 2000 to 2020 for extracted brines and 2015 to 2022 for well samples. Well locations are shown on Figure 1B. In general, brine salinity is highest and most consistent in August and September. Extracted brine density increases in the summer with peak evaporation and decreases in the winter with decreased evaporation and increased precipitation. Halite nucleus salinity reflects the impact of temperature on the solubility of halite. As the crust warms, it’s brine can dissolve more halite, and increase in density. The halite nucleus shallow aquifer sa- linity is highest at the end of summer and lowest in the spring. Lower salinity in early spring may reflect the upward movement of deeper, less saline brine as the surface warms (Bernau and Bowen, 2021). Brine densities in the transitional zone east shallow aquifer generally peak in September, potentially reflecting evapo- concentration. 10 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats and chloride ions exhibit greater densities. When these samples reach halite saturation, as they are when they are in contact with halite for a sustained period, any changes in density are limited by the pro- cess of halite dissolution or crystallization. In the case of brines in contact with halite, there have been in- stances where the reported densities were below the density of halite saturation. This was exceptionally clear in samples from the crust-hosted aquifer with re- ported densities below 1.18 g/cm3, which is below halite saturation (1.195 g/cm3 minimum at halite satu- ration at BSF) (Figure 5). These anomalously high and low measurements were omitted from analyses. The final step in data quality control used known spa- tial distributions of salinity to identify unusually high or low density measurements, which were then as- sessed and removed or kept if they were consistent with additional site measurements during that period. Individual Well-Based Analysis Given the spatial heterogeneity of the brine chem- istry, it is important to evaluate this system at specific sites over time in addition to characterizing the over- all system. Well sites were selected for individual well-based analysis if they had multiple measure- ments across several study periods (a site could in- clude past borings and wells of similar depths at the same location). The Kendall test and linear regression (Kendall package and linear model function in the R© coding language) (Wilkinson and Rogers, 1973a; Hipel and McLeod, 1994; R Core Team, 2021; McLeod, 2022) were applied to site data to identify locations with statistically significant (p-value <0.05) long-term trends in density change. The Kendall test assumes that long-term trends are consistent and have not changed. Because measurements were not uni- formly distributed over time they could not be used for Mann-Kendall or break-point analyses. Areas of Investigation To better describe changes and aggregate data, studied areas were divided spatially and by aquifer depth (Figure 1B and D). Lateral Divisions Areas of investigation were divided based on sur- face mineralogy, potential groundwater flows, fabri- cated structures (the interstate highway and drainage ditches), and lateral salinity (density) gradients (Figures 1B and 4A). Similar terminology based on hydrological fluxes and sedimentology has been used to establish lateral divisions in other saline pans (Munk and others, 2021). From west to east, these ar- eas are the transition zone west (TZW), halite nucle- us, transition zone east (TZE), east of ditches, and I- 80 south (Figure 1). The TZW area includes the region to the west of the persistent halite crust up to the Silver Island Mountain front. This area consists of a mudflat (inner TZW) which transitions into a higher-elevation mud- flat with dunes and intermittent vegetation (outer TZW). The location of the inner and outer TZW is re- flected in USDA soil maps, where the inner TZW ar- ea corresponds with a playa unit and the outer TZW area corresponds with a playa-saltair complex with 0 to 1% slope (Soil Survey Staff, accessed March 2023). TZW waters flowed toward the saline pan in the past. Before 1946, at least two now-dormant springs near the mountain front flowed between 0.02 and 3.4 m3/minute (Utah Division of Water Rights database). Lines (1979) reported a decline in hydraulic gradient between the alluvial fan and the saline pan. Mason and Kipp (1998) also reported outer TZW desiccation fractures (some >1 m wide) and no hydraulic gradient between the saline pan and TZW. Kipnis and Bowen (2018) also noted a decline in alluvial-fan aquifer groundwater levels beyond historical norms after 1998. The halite nucleus consists of an area with a per- sistent halite crust (up to 1.5 m thick). The minimum halite extent, as mapped across several decades of aerial imagery, was interpreted as the halite nucleus’ boundary. The salt crust aquifer only occurs beneath the halite nucleus. Nolan (1927) noted the halite nu- cleus extended south to the area of the current potash mine in 1925. The TZE is between the halite nucleus and the eastern brine extraction ditches. This area is covered by ephemeral halite crust (precipitated from standing water) and efflorescent salts (primarily halite) overly- ing authigenic gypsum sand (Bernau and Bowen, 2021). The area to the east of the brine collection ditches is hydraulically connected to the greater GSLD. The east of ditches area is defined as the re- gion that is closest to the drainage ditches and is like- ly to be impacted by brine extraction. The final area of investigation is the zone immediately to the south of I-80 that is impacted by brine extraction ditches and is isolated from the saline pan by I-80, which pre- vents overland flow and limits subsurface brine movement (Mason and Kipp, 1998). Before manmade structures were built at BSF, brines could flow from the southern and eastern parts of the GSLD to BSF. Evidence of this can be seen today in aerial and satel- lite imagery of seasonal ponds that develop to the 11 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 southeast of the potash mine (Radwin and Bowen, 2021). Aquifer Divisions Previous studies characterized basin-fill, alluvial- fan, and shallow brine aquifers by water chemistry and recharge rates (Turk, 1973; Lines, 1979; Mason and Kipp, 1998). Aquifers and samples are delineated by depth (Figure 1D); these include surficial samples, such as brine collected from extraction ditches and the laydown, and subsurface samples, which are the focus of this study. Depth intervals are described from shallowest to deepest. The crust-hosted wells occur in the halite nucleus, where they are screened within 1 m of the surface. Brine samples from this aq- uifer should be at halite saturation because the aquifer is hosted in halite. The shallow aquifer occurs directly under the crust-hosted aquifer in the halite nucleus and is in contact with the surface elsewhere. The shal- low, moderate, and deep aquifers occur within lacus- trine to saline sediments which consist of carbonate- rich mud and gypsum (with gypsum only occurring in deep basinal muds, >50 m depth) (Shuey, 1971; Ste- phens, 1974; Oviatt and others, 2020; Utah Division of Water Rights database). The shallow basinal mud aquifer (reported as the shallow brine aquifer in other publications) occurs from ≥0 to <10 m depth. This aquifer ranges in salini- ty from 1.04 to 1.21 g/cm3 (Figure 4A), is fractured, and contains brine-shrimp fecal pellet intervals, con- tributing to a higher hydraulic conductivity than an- ticipated from its mean grain size of silty clay (Turk and others, 1973; Lines, 1979). The aquifer’s frac- tures occur in hexagonal patterns. Turk and others (1973) proposed that the fractures formed through os- motic desiccation or synaeresis. Where multiple wells exist in the same aquifer depth range at the same site, the shallower well or the well with a longer reporting span was used for multi-decadal single-well analyses. Turk (1973), Turk and others (1973), and Lines (1979) estimated the total thickness of this aquifer to be between 4.5 and 8 m. The largest source of re- charge to the shallow brine aquifer is meteoric water infiltration through the surface (Mason and Kipp, 1998). Major aquifer discharge sources are the pump- ing of ditches along the eastern margin of the saline pan and subsurface flow south underneath I-80 (Mason and Kipp, 1998). Wells screened in the moderate depth aquifer oc- cur within ~10 to 30 m depth and only occur within the halite nucleus and TZE. Permeability in the mod- erate depth aquifer is far lower than the shallow aqui- fer (Mason and Kipp, 1998), possibly due to the ab- sence of fractures and limited connection with overly- ing higher-permeability aquifers. Alluvial-fan brackick water (BW) aquifer wells (>10 to 150 m depth; measured wells occur between 22 and 111 m depth) are screened in muds (which occur from sur- face to 8 to ~70 m depth across BW wells) to alluvial- fan gravels (Stephens, 1974; Bernau and others, 2023b). The observation well (OW) alluvial-fan aqui- fer wells (Figure 1C) do not have any reported logs with them, but the OW wells closer to the Silver Is- land Mountain front reach depths known to intersect gravel lenses. Two wells at BSF’s center occur within deep (~30 to ~250 m depth) basinal muds and possibly bedded gypsum; this lithology interpretation is based on deep brine well logs from the potash mine to the south (Utah Division of Water Rights database; Bernau and others, 2023b). These deep basinal wells have 3-m- long screens at ~70 and 150 m depth. Underlying deep basinal muds are basinal gravels, which occur at depths of >250 m. The potash mine uses wells in basinal gravels as a source of potassium-rich brine. The basinal gravel aquifer consists of gravels, con- glomerates, and Tertiary volcanic rocks (Stephens, 1974). Water rights reports and data reported in the potash mine reclamation plan (Shaw Environmental, 2020) of basinal gravel wells show the deep brine aq- uifer’s potentiometric surface declined ~20 to 30 m between the 1950s and 2010s. RESULTS Here, chemical results are analyzed in a spatial and temporal context, progressing from west to east, covering the period from 1925 to 2023. Datasets with insufficient information to differentiate trends are not discussed. For example, most trace elements had in- sufficient data to identify spatial or temporal changes. Transitional Zone West Compositionally, many TZW samples differ markedly from other BSF samples (Figure 7). They have higher relative proportions of sulfate (SO4 2-), al- kalinity (as HCO3 -), calcium (Ca2+), and magnesium (Mg2+) than other areas because they have lower con- centrations of sodium (Na+) and chloride (Cl-) (especially in the alluvial-fan aquifer wells). Shallow aquifer TZW brines show a clear decrease in magne- sium between the 1964–1972 and 1999–2006 periods, with increasing magnesium after the 1999–2006 peri- od (Figure S1). Additionally, lithium (Li+) concentra- tions are much lower in the TZW than in other areas (Figure S2). Analysis of individual wells in the TZE shallow 12 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats aquifer identified two areas of increasing salinity (Figures 8A and B). The first segment occurs near the OW transect (Figure 1C). The second segment occurs mid-way along the saline pan, where the inner TZW is wider because of a low-lying Silver Island Moun- tains pass connecting this area to Pilot Valley. Aggregate analysis of the TZW shallow aquifer (Figure 9A) indicates that the outer TZW area has ex- perienced an increase in salinity over time. In con- trast, the inner TZW area does not exhibit consistent changes in salinity. The inner TZW aggregate results contrast with those of the individual well analyses, possibly from differences in sampling location across studies. Over time, the outer TZW area has had marked declines in groundwater levels; several shal- low aquifer wells were persistently dry during the 2013–2023 study period. Stable water isotope measurements from the TZW shallow aquifer indicate that it is isotopically lighter (meaning it originates from less evaporated waters or precipitation from cooler periods) relative to other shallow aquifer areas (Figures 10 and 11). Similar to the shallow aquifer in other areas, a negative shift in TZW shallow aquifer deuterium excess values (Dansgaard, 1964) suggests a change to more evapo- rated waters over time (Figure 11). The alluvial-fan aquifer has had notable declines in groundwater levels since the early 1990s (Figure 3E) (Kipnis and Bowen, 2018). In addition to hydrau- lic head changes, between 1993 and 2022 there have been marked changes in brine density and δ2H and δ18O values. Furthermore, the spatial distribution of these values over time (Figures 12–15) reflects changing groundwater sourcing from mountain front sourced waters to more evaporated waters from the saline pan area. The BW production well transect shows increas- ing density and a shift to heavier δ2H and δ18O iso- topes over time (Figure 12). The largest density in- crease is concentrated at the center of the active pro- duction field (where produced waters now exceed a density of 1.05 g/cm3), with smaller density increases occurring on the edge of the active field (wells at 1.8 and 4.2 km). There are some exceptions to the corre- lation between increased density and generally heavi- er water isotopes. Waters from a well at 4.2 km had relatively low densities, but heavier isotopic values. This indicates that some waters originate from precip- itation under warmer conditions or from evaporated waters (as suggested by water isotopes) that do not have an elevated salinity. Observations from the OW well transect, which spans the inner to outer parts of the alluvial-fan aqui- fer (Figure 1C), inform the interpretation of observed changes in BW chemistry (Figures 13 to 15). Figure 13A to C, a cross section of OW measurements over time shows an increased density gradient towards the basinward direction (to the east) with a transition in salinity between 1 and 2 km. Similarly, Figures 15A and D show a transitional zone between the alluvial- fan and basinal δ2H and δ18O values in the year 1993 occured between 1 and 2 km along the transect. The δ2H, δ18O, and deuterium excess values observed in mountain front-adjacent alluvial-fan samples suggest these waters are sourced from winter precipitation that has undergone minimal evaporation and fraction- ation (Figure 11). The deuterium excess values of all Figure 7. Piper diagram of geochemistry measurements across BSF at A) normal scale, B) magnified scale. A) The high salinity system dominated by Na and Cl makes differ- entiating between sites based on relative ionic content challenging. B) Differentiation between sites when exam- ined at the 90/10/10 percentile values. 13 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 A B Figure 8. A). Select well trends in density through time. Samples from east of ditches, TZE, and I-80 south areas are all from wells screened in the shal- low aquifer. See Figure 1B well locations. B) Sites from Fig. 8A with a statistically significant change in salinity. Color is used to denote change in densi- ty. Size denotes the relative fit of change with time (higher r2 shows a stronger correlation of change with time). All wells in B) are in the shallow aquifer except for the alluvial-fan aquifer well in the south- west corner of the map.  14 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats Figure 9. Box and whisker plots of changes in (A) shallow aquifer brine density across differ- ent areas over time and in (B) brine extraction rates over time. Individual well data has been averaged for each sampling period to minimize the effect of sampling bias on results. Note that sampling has not been consistent at the same wells over time, making trends identified here different from those identified in Figure 8A. Extraction rates before 1991 are primarily esti- mated and have a high range of uncertainty. 15 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 other areas suggest their composition was influenced by evaporation. Nevertheless, the absence of a pro- portional increase in deuterium excess values with in- creasing brine density implies that the salinity of the brine arises not only from evapoconcentration but al- so from the dissolution of salt (Figure 11D). These data strongly support the theory that high salinity in the halite nucleus is maintained by salt crust dissolu- tion. Between 1993 and 2022, the hydraulic gradient shifted to mountain front-directed flow with a 10 m change in the hydraulic head at 0 km in the OW tran- sect (Figure 14A to F). Hydraulic head was corrected for density to assess the effect of density on ground- water flow by using Equation 1 (Figure 14D and E) (Post and others, 2007). (Equation 1) Figure 10. Spatial changes in water δ2H and δ18O. A) All isotope data. B) Average isotope values by area with standard deviation in error bars. TZE alluvial outer wells indicate OW wells located >1 km east of transect start. LMWL is the local meteoric water line. LEL is the local evaporation line. 16 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats where ρ1 is the reference density to adjust the sample to (freshwater density, 1.0 g/cm3); ρ2 is the density of the well-water (calculated from the average density of water’s produced from a well in a period); h2 is the height of the water level above a datum (mean sea level); z is the elevation (above the sea-level datum) of the mid-point of the well’s screened interval; and h1 is the equivalent head relative to the datum. Between the years 1993 and 2022, there was an approximately 12-meter change in hydraulic head at the center of the BW production field. Interestingly, the basal head level observed in the OW wells (Figure 14b) for 2022 closely matches the head observed at the center of the production field in Figure 3E. This similarity in head change, despite a lateral offset of over 4 kilometers, indicates a high level of hydraulic connectivity across the BW alluvial-fan aquifer wells. Correcting for the effect of density on hydraulic head has a significant impact on inferred water flow, as shown in Figure 14. In 1993, it becomes evident that groundwater primarily flows towards the west when the density correction is applied. Without this correction, groundwater flow would have been inter- preted as moving towards an elevation of approxi- Figure 11. Comparison of stable isotopes by area and aquifer over time. A) δ2H, B) δ18O, C) deuterium excess, and D) deuterium excess relative to density. 17 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 mately 1240 meters above sea level at 2 kilometers along the transect. In 2022, the density correction has a relatively smaller impact, but it still highlights that groundwater primarily flows in a westward direction, and downward vertical flow is less significant than what the uncorrected head measurement would sug- gest. Changes in OW-brine density indicate that brine transports salt mass from the halite nucleus crust and underlying aquifers and towards the alluvial fan. Changes are concentrated on the basinward side of the OW cross section; where salinity has increased by 0.04 g/cm3 at 2.6 km at ~1255 meters above sea level and by ~0.01 g/cm3 at ~1215 meters above sea level. The cross-sectional view of salinity change (Figure 13C) shows a saline “nose” with brine migrating down and then west and upwards towards the alluvial fan (possibly along gravel lenses). OW-transect wells have the largest stable-isotope changes of any area in the dataset. The largest change occurs at ~0.8 km at ~1245 meters above sea level (Figure 15C and F). This change reflects the move- ment of basinal water in the direction of the mountain front and is also shown in the 2022 cross-sectional view of isotope data (Figure 15B and E), which shows a blurring of the delineation between alluvial- and basinal-sourced water that was evident in the past (Figure 15A and D). These OW changes suggest that the BW well at 4.2 km has tapped basinal waters but has not yet sourced waters from areas with elevated salinity. Halite Nucleus Figure 5 illustrates declining halite nucleus brine density with depth. Crust aquifer brine is halite satu- rated (>1.195 g/cm3); the shallow aquifer has high sa- linity (1.175 to >1.195 g/cm3), and the moderate and deep aquifers have lower salinities (1.09 to 1.175 g/ cm3 and 1.08 to 1.11 g/cm3, respectively). Mason and others (1995) showed this decline in salinity with depth by measuring pore water chemistry at multiple depths. Individual well plots show no change in density over time in the halite nucleus crust samples (Figure 5). Past research suggested that the crust aquifer po- tassium concentrations decreased between the 1960s and the 1970s (Lines, 1979); subsequent analyses, however, show no long-term change in potassium from the 1960s baseline (Mason and Kipp, 1998; White, 2002) (Figure S3). Calcium concentrations de- creased between the 1960s (from a high of ~1700 mg/ L) to an observed low in the 1991–1997 period (~1100 mg/L) and then increased (Figure S4). There is a notable isotopic lightening in the halite nucleus crust aquifer for δ2H values between the peri- ods of 1991–1997 and 2013–2023. Shallow aquifer samples show a similar change in δ2H. This trend is not seen in the δ18O values, which slightly increased in both aquifers. Two wells showed decreasing halite nucleus shal- low aquifer salinity over time (Figure 8). Aggregate data show a long-term density decrease with brines becoming halite undersaturated between 1964 and 1997 (Figure 9A). Afterward, density remained stable and then increased during the 2013–2023 period. Spa- tial differences in sampling location over time may influence this trend. In contrast to density measure- ments, the halite nucleus shallow aquifer sodium and chloride concentrations appear to consistently de- crease over time (from 105 to 90 g/L and from 180 to 150 mg/L, respectively) (Figures S5 and S6), high- lighting the problematic nature of accurately measur- ing high-salinity brines (Bernau and others, 2023a). Calcium concentrations are positively correlated to changes in density over time in the halite nucleus shallow aquifer while sulfate concentrations are nega- tively correlated to density changes (Figures 9, S4, and S7). Moderate and deep halite nucleus aquifer wells Figure 12. BW well cross section with salinity and water stable isotopes over time. Location of A to A’ is shown on Figure 1C. 18 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats Figure 13. OW cross section densi- ty. Density in (A) 1993, (B) 2022, and (C) change in density between 1993 and 2022. Each blue dot rep- resents the midpoint of a well’s screened interval. The location of B to B’ is shown in Figure 1C. (~40x vertical exaggeration). Elevation is meters above sea level. Local sur- face elevation (not shown) is 1284.9 to 1293.0 m. Note that A & B use the same color scales with only the applicable range for each segment being shown. 19 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 have a limited sampling history over a small spatial extent, and no unequivocal changes in their composi- tion were observed. The moderate depth aquifer had the highest observed sulfate concentrations of any ar- ea and aquifer (Figure S7). Transitional Zone East Two wells showed clear decreases in TZE shal- low aquifer salinity over time (Figure 8). Aggregate TZE shallow aquifer measurements show salinity de- creased through the 1991–1997 period and increased afterward. In contrast to density (and similar to the halite nucleus shallow aquifer), reported sodium and chloride concentrations have decreased over time. Magnesium concentrations at TZE reflect observed density changes. Similarly, TZE shallow aquifer po- tassium concentrations decreased up to the 1991– Figure 14. OW cross section of hydraulic head over time. (A to C) Measured hydraulic head measurements and (D to F) hydraulic head measurements corrected for density. Note the impact of density correction on apparent flow direc- tion (red arrows). Note that A & B, and D & E use the same color scales, respectively, with only the applicable range for each segment being shown. Each blue dot represents the midpoint of a well’s screened interval. The location of B to B’ is shown in Figure 1C. (~40x vertical exaggeration). Local surface elevation (not shown) is 1284.9 to 1293.0 m. Elevation is meters above sea level. 20 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats 1997 period; there were no changes in the following period. Measured δ2H became lighter between the 1991–1997 and 2013–2023 periods (Figure 11A). Extracted Brines Here, changes in brine extracted from the eastern collection ditches are described. NaCl removal from subsurface flow to the south (Mason and Kipp, 1998) is not considered in these values. Sustained brine pro- duction at BSF began in 1939 (Bingham, 1980); since then, the location and volume of brine extraction for potash production at BSF have varied over time (Figure 9B). Before 1966, the Salduro Loop ditch at the center of BSF (Figure 1B) was used to harvest brine; this ditch was likely constructed after sustained production began in 1939, and it is present in 1946 Figure 15. OW cross section of water stable isotopes (δ2H: A to C, and δ18O: D to F) over time. A & B and D & E use the same color scales, respectively, with only the applicable range for each segment being shown. Each blue dot rep- resents the midpoint of a well’s screened interval. The location of B to B’ is shown in Figure 1C. Note the vertical scale differs from Figures 13 and 14. (~40x vertical exaggeration). Local surface elevation (not shown) is 1284.9 to 1293.0 m. Elevation is meters above sea level. 21 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 aerial imagery. Hadzeriga (1964) reports an initial 1% KCl and 21% NaCl weight content of brines used for potash production. The mass of potassium produced annually from the shallow aquifer since 1968 is re- ported in the potash mine’s estimated resources and reserves report (Agapito, 2022). Production from BSF area and the area south of I-80 was lumped together, as such these values can only be used to provide an estimate of brine extraction from BSF over time. The eastern ditches at BSF have been pumped intermit- tently since 1963 (Lines, 1979). An estimated 0.26 Mt NaCl/year was extracted from the eastern ditches be- tween 1966 and 1972 (Stephens, 1974) (Figure 9B). Lines (1979) reports extraction of >0.25 Mt NaCl/ year in 1976. Mason and Kipp (1998) estimated 0.47 Mt NaCl/year extraction. White (2004) reports that brine extraction between 1995 and 1998 was between 0.45 and 0.94 Mt NaCl/year. Brine extraction rates between 2001 and 2021 were highly variable (Todd Marks, Bureau of Land Management, written commu- nication, 2023), ranging from almost none to nearly 0.8 Mt of NaCl extracted/year (Figure 3A). There was over twice as much annual average extracted mass in the 1996–2006 period (0.40 Mt/year) than in the 2013 –2023 period (0.16 Mt/year). During at least the last part of the 1996–2023 period, the northern part of the extraction ditches was inactive (Figure 1B; potash mine personnel, verbal communication, 2022). Exam- ination of the ratio of total KCl produced from the shallow aquifer (BSF and area south of I-80) and NaCl extracted from BSF over time shows that pro- duction became less reliant on BSF after the year 2000; with an average of 10 tonnes of NaCl extracted from BSF for every ton of KCl produced by the mine before 2000, and 5.5 tonnes of NaCl extracted from BSF for every ton of KCl produced afterward. While seasonal precipitation and evaporation changes can impact the salinity of extracted brine (Figure 6), month-to-month analysis of density and extracted brine volumes show a clear decrease in ex- tracted brine salinity with increased extraction rates (Figure 3D). The decrease in salinity with increased extraction suggests that deeper, less-saline waters rise upward (with reduced hydrostatic pressure) and con- tribute to the shallow aquifer. From 2000 to 2004, the density of the extracted brine increased while the volume of brine extracted remained constant. From 2004 to 2015 the extracted brine density decreased; from 2004 to 2010 there was no net addition of NaCl to the salt flats. Following 2010, brine extraction greatly decreased, while laydown increased, leading to increased net volumes of brine contributed to BSF. Apparently, a 5-year lag occurred between decreased extraction with increased net brine contributions and the onset of salinity in- crease within the system. This is a much longer peri- od than the salinity recovery period in the early 2000s, which was associated with a much larger pulse of input solutes over a shorter period. Following in- creasing density up to 2015, density remained high during the 2016–2020 period. East of Ditches and I-80 South Both the east of ditches and I-80 south areas have reduced hydraulic connection with the halite nucleus and are isolated from laydown brines. Furthermore, brine extraction ditches impact both of these areas. Accordingly, these areas provide an example of how the brine system may respond to extraction on deca- dal timescales without external solute sources. In contrast to the halite nucleus and TZE shallow aquifers, brine samples from the east of ditches and I- 80 south shallow aquifers show decreases in density across individual wells and in aggregate (Figures 8A and 9). Similar to the halite nucleus crust aquifer and TZE shallow aquifer, calcium concentrations in the east of ditches shallow aquifer decreased until the 1991–1997 period (from ~1500 to 1100 mg/L) and in- creased afterward (to ~1400 mg/L) (Figure S4). DISCUSSION Brine Chemistry Changes Over Time BSF brine chemistry changes lie within three groups: 1) no change, 2) long-term decrease or in- crease, and 3) change in long-term trend following the 1991–1997 period. No long-term changes in the deep and moderate-depth aquifers underlying the hal- ite nucleus and TZE were observed. Similarly, the in- ner TZW area, when taken in aggregate, did not show any change in density over time. However, analyses of several wells from the TZW area indicate parts of this area increased in density over time. The shallow aquifer in the east of ditches and I-80 south areas show clear decreases in density over time, in contrast, the outer TZW area is the only shallow brine area to show long-term increases in density across several studies. The BW production well area in the TZW also shows consistent long-term increases in alluvial-fan aquifer density, but the east of ditches and I-80 south areas show long-term decreases in brine density, with the I-80 south area showing the largest salinity decrease. The changes in brine salinity in the I-80 south area also suggest that there is limited transport of solutes under I-80 from BSF to the I-80 south area. High connectivity between these areas would likely limit salinity decreases in the I-80 south 22 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats area as transported salt would replenish removed so- lutes. The last group of chemical change, with changes in long-term trends after the 1991–1997 period, indi- cates a change in saline pan conditions. Long-term density decreases ceased and salinity increased after the 1991–1997 period in the halite nucleus and TZE shallow aquifer (Figure 9A). On smaller timescales, brine extracted for potash production shows an in- crease in salinity following the onset of the laydown in 1998. Later, an increase in salinity occurred when brine extraction greatly decreased as the laydown continued (Figure 3A to C). The onset of the laydown coincides with a marked decrease in alluvial-aquifer groundwater levels. De- clining levels, in turn, reversed hydraulic gradients, enabling basinal brine movement away from the sa- line pan (as seen in density and isotopic changes) (Figure 16). Careful consideration of the underlying forces driving changes in brine chemistry, primarily density and water stable isotopes (e.g., laydown, de- crease in brine extraction, or long-term groundwater consumption trends), is needed to identify the core controls on change that will influence management decisions. Potential Controls on Brine Chemistry Change Laydown A cessation in long-term density decreases and an increase in density in the halite nucleus and TZE shal- low aquifers that is concurrent with the laydown was observed. Similarly, extracted brines for potash pro- duction showed increased density in the 3- to 5-year period after the laydown began. These observations support White’s (2004) hypothesis that the laydown failed to lead to a 4–5 cm increase in halite thickness because it buffered salinity decreases in the shallow brine aquifer. Although increases in aquifer density are concurrent with the onset of the laydown, chang- ing potash brine extraction rates may have had a larg- er impact on observed changes in groundwater densi- ty than the laydown. An δ2H isotopic lightening in both halite nucleus crust brines and shallow aquifer waters over time was observed; in the absence of other information, this lightening could be attributed to isotopically lighter laydown waters. However, isotopic measurements from the east of ditches area (which is isolated from Figure 16. Conceptual model for pre-alteration and modern groundwater (and brine) movement at BSF shown in cross section (left) and as a simplified conceptual model (right). 23 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 laydown waters) show the same lightening over time. Therefore, a different mechanism, possibly increased infiltration of meteoric precipitation, may explain ob- served changes. Changing Brine Extraction Rates BSF brine extraction volumes before 1995 are poorly reported, leading to high uncertainty in esti- mated extraction volumes (Figure 9B). The average extraction volumes during the 1964–1972 and 1975– 1981 periods may have been similar to or higher than those reported between 1995 and 1998. Measure- ments from 1999 to 2023 show stabilizing to increas- ing density in the halite nucleus and TZE shallow aq- uifers. There is a corresponding decrease in brine ex- traction during this period, with the largest increase in the halite nucleus shallow aquifer salinity correspond- ing to the greatest decrease in extraction rates. This relationship suggests that reduced extraction rates may contribute to some (and possibly most) of the ob- served density increases in these areas. Examination of increasing sulfate concentrations over time relative to decreasing density in the halite nucleus and I-80 south areas shallow aquifers indi- cates these two values are related (Figures 9 and S7). This relationship suggests a mechanism for brine re- placement after extraction. The moderate depth aqui- fer has elevated sulfate concentrations. When ground- water is extracted for mining, it lowers the constrain- ing hydrostatic pressure, enabling deeper, less saline groundwater with higher sulfate concentrations to rise and replace extracted waters. There is a positive correlation between extracted brine density and potash production brine extraction volumes (Figure 3D). Using that relationship, antici- pated changes in density based on extracted brine vol- umes over time were modeled. While not reflecting observed density measurements (modeled density of 1.10 to 1.15 g/cm3 between 2002 and 2010, when ob- served density was ~1.17 to 1.21 g/cm3), modeled density did replicate trends in brine density change for some periods. The model replicates observed trends between 2002 and 2006 where it shows in- creasing and then decreasing density; it also shows in- creasing and then stabilized density between 2015 and 2021. Modeled and observed trends strongly dif- fer between 2006 and 2015, where the model shows generally increasing density, while observations show generally decreasing density. These differences be- tween observations and modeled changes suggest 1) extracted brine density may respond non-linearly to extraction rates (and accordingly, there is a lower lim- it for extracted brine salinity, potentially buffered from the dissolution of the halite crust); and possibly 2) under normal operating conditions (with some ex- traction and laydown volume of ~0.5 Mt NaCl/year) extracted brine salinity (and aquifer salinity) will de- crease; additionally, 3) the large initial laydown pulse between 1998 and 2000 increased the salinity of ex- tracted brines beyond their anticipated baseline. Brine extraction can also depress local groundwa- ter levels (Turk, 1973). These declines may enable in- filtration of surface waters into the subsurface before they can evaporate. Isotopic lightening of the shallow aquifer (Figure 11A) may be attributed to the incorpo- ration of isotopically lighter winter precipitation into the aquifer before significant evaporation occurred. Given these observations, especially that changes in salinity can be attributed to extraction rates, the rel- ative role of the laydown in increasing TZE and the halite nucleus shallow aquifer salinity remains un- clear. It may be that a solute source (such as a halite crust or laydown brines) is necessary for density val- ues to recover. Recent work on the sedimentology of BSF salt crusts documented extensive evidence of halite dissolution, suggesting that the diminishing crust is a likely source of these solutes (Bernau and Bowen, 2021). Declining TZW Groundwater Levels Two areas of increasing salinity were identified in the inner TZW shallow aquifer, and a long-term ag- gregate trend in increasing salinity in the outer TZW shallow aquifer was also found. Kipnis (2021) and Lines (1979) noted decreasing groundwater levels in these areas. Furthermore, several dry outer TZW shal- low wells were noted during this investigation. Two mechanisms for increasing TZW shallow aquifer sa- linity are proposed: 1) the TZW inner area receives brine from the halite nucleus aquifers and surface pre- cipitation, and 2) salinity that was once concentrated at the surface by efflorescence and then recycled basinward now accumulates in groundwater as the ca- pillary fringe falls below the ground surface. Flow of halite nucleus brine to the inner TZW area may be en- hanced by declining TZW hydraulic head. Further- more, there is the potential for laydown brines (which accumulate on the western halite nucleus edge) to en- ter the TZW area. Several observations, such as declining western halite nucleus shallow aquifer salinity and areas of in- ner TZW with increasing salinity, support the inter- pretation that westward movement of shallow aquifer brine contributes to recent declines in BSF crust thickness. Between the 2003 and 2016 salt crust thickness studies, the area on the southwestern side of BSF had some of the largest observed decreases in salt crust thickness. This area is also the closest part 24 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats of BSF to the alluvial-fan production wells. During the same period, the northernmost parts of the BSF crust reduced in volume; those declines may be at- tributable to their distance from the thickest part of the halite crust. The alluvial-fan aquifer shows several significant changes over time. The decline in groundwater levels has reversed the hydrological gradient such that basi- nal waters now flow toward the mountain front. The movement of basinal waters toward the mountain front is demarcated by changes in brine density and δ2H and δ18O isotopes. Formerly fresh production wells now produce waters that exceed the salinity of the ocean (~35 ppt or 1.03 g/cm3). These changes in- dicate that basinal brine (possibly the halite nucleus shallow and crust aquifer brine) is being removed from the saline pan area (Figure 16). Isotopic and density data should be considered to determine the relative sourcing of waters extracted by production wells to estimate how much brine will be removed from the halite nucleus area under different alluvial- fan extraction scenarios. Climate Does Not Explain Observed Alluvial- Fan Groundwater Level Declines Between 1993 and 2010, there were frequent drought periods, with the majority of years experienc- ing precipitation levels below the 25% quartile for precipitation based on data from 1910 to 2020 (Bernau, 2022). Before the 1990s, alluvial-fan aquifer levels varied but regularly returned to the land sur- face. In the period since then, they have shown a clear long-term decline that strongly differs from observa- tions of similar alluvial-fan aquifer wells in the GSLD, which have remained relatively stable to slightly increasing over this period (NWIS, https:// maps.waterdata.usgs.gov/mapper/index.html, sites 404757112582701 and 394905113354101). These da- ta indicate that recent declines in the alluvial-fan aq- uifer are occurring because extraction rates have ex- ceeded recharge rates for over two decades. Impact of Laydown on Alluvial-Fan Aquifer Drawdown Before 1997, the alluvial-fan groundwater level remained within 0 to 7 m of the surface (Kipnis and Bowen, 2018; Mason and Kipp, 1998). Following the laydown, groundwater levels never rose above 10 m below ground level and have continued to decline (Figure 3E). This suggests that groundwater levels have not yet reached a new equilibrium where inflow is equal to pumping rates and a larger area will be drained over time. The relative proportion of basinal water in produced alluvial aquifer waters will in- crease as this area expands. The laydown has led to as much as a doubling in alluvial groundwater extrac- tion, and as such, plays a major role in decreasing al- luvial-fan groundwater levels. CONCLUSIONS New chemical and groundwater level measure- ments and past research were compiled to examine multi-decadal changes at the Bonneville Salt Flats. Brine chemistry, most notably density and δ2H and δ18O water isotopes, has changed in response to an- thropogenic activities (Figure 16). Shallow aquifer brine under and to the east of the crust declined in sa- linity between 1964 and 1997 and stabilized and in- creased in salinity afterward. Increased salinity may be due to decreased extraction rates in the past two decades, especially as the largest increase in salinity, during the 2013–2023 period, is concurrent with the largest decrease in extraction. However, this period is also concurrent with the experimental salt restoration laydown program. The relative role of the laydown in increasing aquifer salinity remains unclear. Alluvial aquifer groundwater levels have declined over time. This decline is linked to industrial water production, including the laydown. As a result, the hydraulic gra- dient has reversed, causing brine to flow away from the saline pan and towards the alluvial aquifer. This flow increases alluvial fan aquifer salinity and chang- es its isotopic composition. If alluvial-fan extraction rates remain the same, or if they rise with increases to the laydown, more salt will be removed from the Bonneville Salt Flats halite nucleus, potentially at volumes exceeding the laydown. These multi-decadal chemical changes inform the understanding of groundwater movement and halite crust changes in this system, which informs management for the sus- tained use of this landscape. Supplemental Data Supplemental data and figures are available at https://doi.org/10.5281/zenodo.8152647. ACKNOWLEDGMENTS The authors acknowledge that this study was con- ducted on traditionally Newe/Western Shoshone and Goshute lands. The authors would like to acknowledge Craig Peterson and Russ Draper with Intrepid Potash; Todd Marks, Roxanne Tea, Steve Al- len and other past and current U.S. Bureau of Land 25 M.D. Vanden Berg, R. Ford, C. Frantz, H. Hurlow, K. Gunderson, G. Atwood, editors 2024 Utah Geological Association Publication 51 Management West Desert Field Office staff; Hannah Stinson, Lily Wetterlin, Olivia Watkins, Candace Penrod, Gabë Regenhardt, and Dr. Anna Cassell from the University of Utah; and Paul Inkenbrandt, Hugh Hurlow, and Greg Gavin of the Utah Geological Sur- vey for assistance with data collection and sample analysis. This manuscript was improved through re- views and discussion with Greg Carling, Stephanie Carney, Scott Hynek, Michael Hylland, and Elliot Jagniecki. Torrie Duncan and John Good contributed to refining some of the figures in this report. Cryo- genic vacuum extraction of brine samples was made possible by Suvankar Chakraborty, Dr. Jim Ehlerin- ger, and Dr. Gabe Bowen at the University of Utah SIRFER Lab. Bill White provided unpublished field notes contributing to this work. Funding for this work was provided by an NSF Coupled Natural Human Systems Award #1617473, the United States Geologi- cal Survey, the Utah Geological Survey, the Utah State Legislature, the Bureau of Land Management, and the University of Utah Global Change and Sus- tainability Center Graduate Research Grants. REFERENCES Agapito Associates, Inc., 2022, Technical Report Summary of 2021 Estimated Resources and Re- serves at Intrepid Potash-Wendover, 262 p. Bernau, J.A., 2022, Spatial and temporal scales of water and salt movement at the Bonneville Salt Flats: Salt Lake City, University of Utah, Ph.D. dissertation, 195 p. Bernau, J.A., and Bowen, B.B., 2021, Depositional and early diagenetic characteristics of modern sa- line pan deposits at the Bonneville Salt Flats, Utah, USA: Sedimentology, p. sed.12861. Bernau, J.A., Jagniecki, E.A., Kipnis, E.L., and Bow- en, B.B., 2023a, Applications and limitations of portable density meter measurements of Na-Ca- Mg-K-Cl-SO4 brines: Chemical Geology, v. 616, p.121240 Bernau, J.A., Oviatt, C.G., Clark, D.L. and Bowen, B.B., 2023b, Sediment logs compiled from the Great Salt Lake Desert, western Utah, with a fo- cus on the Bonneville Salt Flats area: Utah Geo- logical Survey Open-File Report 754, 24 p., 3 ap- pendices, https://doi.org/10.34191/OFR-754. Bethke, C.M., 2013, Geochemical and biogeochemi- cal reaction modeling: Cambridge University Press, Cambridge, UK, p.1689–1699. Bingham, C.P., 1980, Solar production of potash from the brines of the Bonneville Salt Flats: Utah Geo- logical and Mineral Survey Bulletin, v. 116. Bowen, B.B., Kipnis, E.L., and Pechmann, J.M., 2018, Observations of salt crust thickness change at the Bonneville Salt Flats from 2003-2016, in Emerman, S.H., Bowen, B.B., Simmons, S., and Schamel, S. editors, Geofluids of Utah: Utah Geo- logical Association Publication 47, p. 247–285. Bowen, B.B., Kipnis, E.L., and Raming, L.W., 2017, Temporal dynamics of flooding, evaporation, and desiccation cycles and observations of salt crust area change at the Bonneville Salt Flats, Utah: Geomorphology, v. 299, p. 1–11. Dansgaard, W., 1964, Stable isotopes in precipitation: Tellus, v. 16, no. 4, p. 436–468. Hadzeriga, P., 1964, Some aspects of the physical chemistry of potash recovery by solar evaporation of brines: Society of Mining Engineers, v. 013, p. 169–174. Harvie, C.E., Moller, N., and Weare, J.H., 1980, The prediction of mineral solubilities in natural wa- ters—the Na-K-Mg-Ca-Cl-SO4-H2O system from zero to high concentration at 25°C: Geochimica et Cosmochimica Acta, v. 44, no. 7, p. 981–997. Hipel, K.W., and McLeod, A.I., 1994, Time series modelling of water resources and environmental systems: Elsevier Science, Amsterdam, Nether- lands. Kipnis, E.L., 2021, Geologic change, hydrologic driv- ers, and resource use at the Bonneville Salt Flats, Utah, USA: Salt Lake City, University of Utah, Ph.D. dissertation, 100 p., 2 appendices. Kipnis, E.L., and Bowen, B.B., 2018, Observations of salt crust change from 1960-2016 and the role of humans as geologic agents at the Bonneville Salt Flats, Utah, in Emerman, S.H., Bowen, B.B., Sim- mons, S., and Schamel, S. editors, Geofluids of Utah: Utah Geological Association Publication 47, p. 287–303. Kipnis, E.L., Bowen, B.B., Hutchings, S.J., Hynek, S.A., and Benison, K.C., 2020, Major ion geo- chemistry in Na-Ca-Mg-K-Cl-SO4 brines using portable X-ray fluorescence spectrometry: Chemi- cal Geology, v. 558, no. September, p. 119865. Lerback, J.C., Hynek, S.A., Bowen, B.B., Bradbury, C.D., Solomon, D.K., and Fernandez, D.P., 2019, Springwater provenance and flowpath evaluation in Blue Lake, Bonneville basin, Utah: Chemical Geology, v. 529, no. April, p. 119280. Lindenburg, G.J., 1974, Factors contributing to the variance in the brines of the Great Salt Lake De- sert and the Great Salt Lake: Salt Lake City, Uni- versity of Utah, M.S. thesis, 70 p. Lines, G.C., 1978, Selected ground-water data, Bonneville Salt Flats and Pilot Valley, western Utah: U.S. Geological Survey Utah Basin-Data Release, no. 30, 14 p. Lines, G.C., 1979, Hydrology and surface morpholo- gy of the Bonneville Salt Flats and Pilot Valley playa, Utah: U.S. Geological Survey Geological 26 J.A. Bernau, B.B. Bowen, E.L. Kipnis, and J.C. Lerback Observations of Decadal-Scale Brine Chemistry Change at the Bonneville Salt Flats Survey Water-Supply, v. 2057, no. 2057, p. 1– 107. Mason, J.L., Brothers, W.C., Gerner, L.J., and Muir, P.S., 1995, Selected hydrologic data for the Bonneville Salt Flats and Pilot Valley, western Utah, 1991-93: U.S. Geological Survey Open-File Report, v. 95–104, p. 1–56. Mason, J.L., and Kipp, K.L., 1998, Hydrology of the Bonneville Salt Flats, northwestern Utah, and simulation of ground-water flow and solute transport in the shallow-brine aquifer: U.S. Geo- logical Survey Professional Paper, v. 1585, p. 108. McLeod, A.I., 2022, Kendall Rank Correlation and Mann-Kendall Trend Test. R code Munk, L.A., Boutt, D.F., Moran, B.J., McKnight, S. V., and Jenckes, J., 2021, Hydrogeologic and geo- chemical distinctions in freshwater‐brine systems of an Andean salar: Geochemistry, Geophysics, Geosystems. v. 22, no. 3, 20 p. https:// doi.org/10.1029/2020GC009345 Nolan, T.B., 1927, Potash brines in the Great Salt Lake Desert, Utah, Chapter B, in Contributions to economic geology (short papers and preliminary reports): U.S. Government Printing Office, p. 25– 44. Oviatt, C.G., Clark, D.L., Bernau, J.A., and Bowen, B.B., 2020, Data on the surficial deposits of the Great Salt Lake Desert, Bonneville Salt Flats and east part of the Wendover 30’ x 60’ quadrangles, Tooele County, Utah: Utah Geological Survey Open-File Report, 70 p. Penrod, C.L., 2016, The geochemistry of meteoric water and surface-pond brines at Bonneville Salt Flats, UT: Salt Lake City, University of Utah, M.S. thesis, 30 p. Pitzer, K.S., 1973, Thermodynamics of electrolytes– I—Theoretical basis and general equations: Jour- nal of Physical Chemistry, v. 77, no. 2, p. 268– 277. Plummer, L.N., Parkhurst, D.L., Fleming, G.W., and Dunkle, S.A, 1988, A computer program incorpo- rating Pitzer’s equations for calculation of geo- chemical reactions in brines: U.S. Geological Sur- vey, Water-Resources Investigations Report, v. 88 –4153, p. 707–720. Post, V., Kooi, H., and Simmons, C., 2007, Using hy- draulic head measurements in variable-density ground water flow analyses: Ground Water, v. 45, no. 6, p. 664–671. R Core Team, 2021, R: A language and environment for statistical computing, R Foundation for Statis- tical Computing, Vienna, Austria. https://www.r- project.org/ Radwin, M.H., and Bowen, B.B., 2021, Mapping mineralogy in evaporite basins through time using multispectral Landsat data—Examples from the Bonneville basin, Utah, USA: Earth Surface Pro- cesses and Landforms, v. 46, no. 6, p. 1160–1176. Read, E.K., Carr, L., De Cicco, L., Dugan, H.A., Hanson, P.C., Hart, J.A., Kreft, J., Read, J.S., and Winslow, L.A., 2017, Water quality data for na- tional-scale aquatic research: The Water Quality Portal, Water Resources Research, v. 53, no. 2, p. 1735–1745. Shaw Environmental, 2020, Intrepid Potash Mine and Reclamation Plan, submitted to Utah Division of Oil, Gas, and Mining: 1943 p., https://ut-dnr-ogm- prod-sf-public- bucket.s3.amazonaws.com/5624847.pdf Shuey, R.T., 1971, Paleomagnetic chronology and correlation of Great Salt Lake basin sediments: Washington D.C., National Science Foundation, Final Technical Report for grant GA-16134, 15 p. Soil Survey Staff, Natural Resources Conservation Service, United States Department of Agriculture, undated, Web Soil Survey. Available online: https://websoilsurvey.nrcs.usda.gov/app/ WebSoilSurvey.aspx, accessed March 2023. Stephens, J.C., 1974, Hydrologic reconnaissance of the northern Great Salt Lake Desert and summary hydrologic reconnaissance of northwestern Utah: United States Geological Survey, 55 p. National Water Quality Monitoring Council, undated, The Water Quality Portal, https:// www.waterqualitydata.us/, accessed November, 2022. Turk, L.J., 1973, Hydrogeology of the Bonneville Salt Flats, Utah: Utah Geological and Mineral Survey, 82 p. Turk, L.J., Davis, S.N., and Bingham, C.P., 1973, Hy- drogeology of lacustrine sediments, Bonneville Salt Flats, Utah: Economic Geology, v. 68, no. 1, p. 65–78. Utah Division of Natural Resources, undated, Water Rights database: https:// www.waterrights.utah.gov/wrinfo/query.asp, ac- cessed November, 2022. White, W.W., 2002, Salt Laydown Project— Replenishment of Salt to the Bonneville Salt Flats, in Gwynn, J.W., editor, Great Salt Lake: An Overview of Change: Utah Geological Survey Special Publication, p. 433–486. Wilkinson, G.N., and Rogers, C.E., 1973, Symbolic description of factorial models for analysis of var- iance: Journal of the Royal Statistical Society, Se- ries C, v. 22, no. 3, p. 392–399.