Understanding Groundwater Resource Vulnerability at Bryce Canyon National Park, Utah, Using Applied Geophysics at the Rubys Inn Thrust Fault

Groundwater resource development, the threat of future drought, and long-term effects of climate change in Garfield County, south-central Utah, have prompted some recent studies of groundwater in Bryce Canyon National Park and adjacent areas [1, 2]. Emery Valley, situated predominantly north of Bryce Canyon National Park, is subject to a potential increase in growth from tourism-related development (Figure 1,1). Understanding of the hydrogeology of the region has recently improved but is still incomplete; it is currently uncertain whether groundwater withdrawals in Emery Valley could eventually affect groundwater resources up-gradient throughout the Paunsaugunt Plateau to the south or throughout Bryce Canyon National Park to the south and southeast.

The Rubys Inn thrust fault, referred to hereafter as the Rubys Inn fault, is an east-west-striking thrust fault that lies between commonly targeted aquifers in Emery Valley and groundwater flow systems of the Paunsaugunt Plateau, which support springs, seeps, and dependent ecosystems to the south. We designed this study to improve the understanding of how the Rubys Inn fault influences groundwater movement between the Paunsaugunt Plateau and Emery Valley. This research allows government agencies and land managers to make science-based resource management and water rights-related decisions.

The goal of this study is to improve the understanding of the hydrogeologic control, structural characteristics, and adjacent stratigraphy of the Rubys Inn fault. Fault zones are understood to have heterogeneous permeability and anisotropy, meaning faults can act as both barriers and conduits to flow, depending on their internal structure, geometry, and the geologic units they juxtapose [3, 4]. We describe the survey results for three electrical resistivity tomography (ERT) transects positioned across the Rubys Inn fault near Bryce Canyon City. We also use groundwater-level data and literature- and field-based outcrop observations of the fault to assist with the interpretations of ERT inversion models. These data help determine whether the Rubys Inn fault acts as a barrier or conduit to cross-fault groundwater flow and therefore the likelihood that Emery Valley development might impact groundwater-dependent resources across the fault. The results also provide critical information for designing and identifying suitable locations for future monitoring wells that would enable further investigation of the fault’s hydrogeologic control and long-term groundwater monitoring.

The study area is located near Bryce Canyon City, Garfield County, Utah (Figure 1,1). The area includes the southeast corner of Emery Valley; parts of Bryce Canyon City, a small town adjacent to Bryce Canyon National Park; and the northwest corner of Bryce Canyon National Park itself.

Emery Valley is in the High Plateaus subprovince of the Colorado Plateau, which acts as a transition zone between the Basin and Range province to the west and Colorado Plateau province to the east [5]. This structural and stratigraphic transition zone is defined by a series of north–south-striking, west-side-down oblique-slip normal faults.

Emery Valley is a shallow intermontane basin bounded by the Sevier Plateau to the north and the Paunsaugunt Plateau to the south. The Paunsaugunt Plateau is the main recharge area for both Emery Valley and Bryce Canyon National Park, supplying surface water to the East Fork Sevier River and groundwater to Emery Valley and Bryce Canyon National Park.

The stratigraphy in the study area consists of Upper Cretaceous and Paleocene to Eocene sedimentary rocks, and Quaternary surficial deposits (Figure 2,2). The John Henry Member of the Straight Cliffs Formation (Upper Cretaceous) consists of mudstone and fine-grained subarkosic sandstone, about 240–335 m thick in Bryce Canyon National Park [6]. The Drip Tank Member of the Straight Cliffs Formation (Upper Cretaceous) consists of fine- to medium-grained

quartzose sandstone and pebbly sandstone/conglomerate that ranges from 30 to 60 m thick on the Paunsaugunt Plateau [7]. The Wahweap Formation (Upper Cretaceous) consists of mudstone, fine-grained sandstone, and silty sandstone that thickens east to west on the Paunsaugunt Plateau from around 90 to 210 m, respectively [8]. The Kaiparowits Formation (Upper Cretaceous) consists of fine-grained sandstone, mudstone, and siltstone and is generally locally eroded along the eastern section of the Rubys Inn fault. The pink member of the Claron Formation (Paleocene to Eocene) consists of micritic limestone, calcite-cemented sandstone, and calcareous mudstone and is around 180 m thick in Bryce Canyon National Park [7]. The white member of the Claron Formation (Eocene) typically consists of white micritic limestone which grades to interbedded mudstone, siltstone, and limestone in the northern part of Bryce Canyon National Park [7]. Changes in lithology within the Claron can be attributed to both depositional environment and bioturbation, the latter of which can cause dramatic lateral changes [9]. The conglomerate at Boat Mesa (Eocene) consists of pebbly conglomerate, calcareous and conglomeratic sandstone, and minor siltstone and mudstone that ranges up to 30 m thick in Bryce Canyon National Park [6]. Quaternary unconsolidated deposits in the study area consist of colluvium and alluvium composed of poorly to moderately sorted clay- to boulder-size sediments deposited by slope wash, debris flow, soil creep, and ephemeral fluvial processes [7].

The Rubys Inn fault is part of the Paunsaugunt thrust fault system, a grouping of thrust faults that are middle Tertiary in age that strike perpendicular to the predominant Sevier and Laramide faulting in the region. This fault system resulted from the Markagunt gravity slide, formed during Tertiary time by the gravitational collapse of the southern portion of the Marysvale volcanic field [7, 10].

The Rubys Inn fault is the most prominent of several south-directed, shallow thrust faults with a total offset of up to 460–910 m [5, 11]. The Rubys Inn fault strikes westward from the Paunsaugunt fault zone, where the surface trace is lost at the western edge of the Paunsaugunt Plateau. However, recent research suggests that the fault continues into the Markagunt Plateau to the west as a blind thrust [7]. In the study area, the fault places the John Henry Member of the Straight Cliffs Formation over the pink and white members of the Claron Formation. To the east, the fault places Wahweap Formation over the white Claron, and continuing east, pink Claron on top of itself.

Two recent studies have examined groundwater in Emery Valley in detail [1, 2]. Wallace et al. [2] conducted a comprehensive study of the aquifers of Emery Valley and neighboring Johns Valley to the northeast, as well as groundwater to the south on the Paunsaugunt Plateau and in Bryce Canyon National Park. Their study includes a hydrogeologic framework, a groundwater age analysis, and a water budget. This current study was motivated by their observation that faults, primarily the Rubys Inn fault and Paunsaugunt fault zone, likely play a role in the groundwater system.

The Rubys Inn fault was identified in the mid-1950s but remained undocumented until 1986 and has since been studied over the following four decades [7]. Davis and Krantz [12] discovered evidence of the Paunsaugunt thrust fault system in Bryce Canyon National Park. Lundin [11] mapped the Rubys Inn fault in detail. Nickelsen and Merle [13], Nickelsen et al. [14], Davis [5], Leavitt et al. [15], May et al. [16], and Cleveland et al. [17] have extensively documented thrust faulting and associated folding in and around the Rubys Inn fault. Although the Rubys Inn fault has been studied extensively, no studies have truly addressed groundwater flow relative to the fault.

We used geophysical surveys, hydrogeologic data, and geologic observations to better understand the hydrogeology and structural variability of the Rubys Inn fault. We applied ERT to collect subsurface data, create inversion models, and facilitate structural and hydrogeologic interpretations. This geophysical method provides high-resolution data to resolve complex structure and stratigraphy at varying depths. ERT surveys were conducted in May 2022 and September 2022, the timing of which was intended to capture seasonal hydrologic conditions reflecting presumed maximum and minimum groundwater levels, respectively. We collected water-level data, used existing well logs, and compiled other geologic information to assist in interpretation of geophysical modeling results. We compiled lithologic logs and historical water-level data from six monitoring wells in the study area [18, 19]. We measured water levels from five of these wells (excluding well 66W) in May, June, and September 2022.

ERT is an active-source, subsurface imaging technique that uses a series of electrodes to produce currents and measure electrical potentials along user-designed transects. These measured electrical potentials, or voltage differences, depend on the distribution of electrical resistivity in the subsurface and the geometry of the electrode layout. For a detailed description of ERT methodologies, refer to Johnson et al. [20], Singha et al. [21], and Binley and Slater [22]. Electrical resistivity in the subsurface varies with lithologic and hydrologic conditions, including degree of saturation, porosity, fracturing, pore fluid salinity, and mineral composition [23-26]. For example, highly porous and saturated materials, materials wetted with high-salinity fluids, and materials composed of clay- or metal-rich materials will better conduct the flow of electric current and therefore exhibit lower electrical resistivities than unsaturated materials with low porosity and relativity high silica or carbonate content. This method leverages the contrasting resistivities of different geologic materials to provide a modeled image of subsurface lithology and pore fluid distribution. ERT has been widely used for investigating faults and fault-zone hydrogeology for the past two decades [27-32]. Ball et al. [30] used ERT to supplement hydrologic data to infer increased fracturing in a damage zone surrounding a reverse fault, as well as interpret both a high permeability damage zone (conduit) and low permeability fault core (barrier). Saribudak and Hawkins [31] used ERT and other geophysical methods to demonstrate a fault plane conduit due to karstification at a fault that was previously characterized as a barrier to lateral groundwater flow. Barnes et al. [32] applied ERT alongside geologic and hydrogeologic data to delineate subsurface fault-zone cementation (i.e., a barrier to flow) in Quaternary unconsolidated sediments.

We conducted all ERT surveys with an Advanced Geosciences Incorporated (AGI) SuperSting Wi-Fi R8 electrical resistivity meter and 56-electrode passive cabling system. We surveyed three two-dimensional (2D) ERT transects centered on and roughly orthogonal to the Rubys Inn fault as mapped by Biek et al. [7] from May 17 to May 19, 2022: one 550-m transect with 10-m electrode spacing (BRCA1) and two 220-m transects with 4-m electrode spacing (BRCA3 and BRCA5) (Figure 1,1). From September 13 to September 15, 2022, we repeated the 10-m electrode spacing survey (BRCA1B) and central 4-m electrode spacing survey (BRCA3B) and performed a higher resolution 332-m length 4-m electrode spacing survey along a portion of BRCA1 (BRCA7). Survey BRCA7 was collected “roll-along” style, meaning overlapping data were collected to generate a longer survey line using a shorter electrode spacing to achieve higher resolution over a greater area. Each survey was executed with a hybrid dipole-dipole/strong-gradient array developed and recommended by AGI. The dipole-dipole component provides high-resolution data whereas the strong-gradient component provides higher signal levels. For the ERT method, horizontal and vertical resolution is inversely related to electrode spacing. Maximum resolution is one-half of the electrode spacing. Therefore, surveys designed with 10-m electrode spacing (e.g. BRCA1) can resolve features no smaller than 5 m, whereas surveys designed with 4-m electrode spacing (e.g. BRCA7) can resolve features no smaller than 2 m.

Maximum depth of investigation (DOI) is approximately 25% of a given total array length but depends on the distribution of subsurface resistivities and user-selected electrode spacing, with a longer total array length enabling a greater DOI [33]. Actual DOI for the surveys ranged from 17% to 24% of total array length. We measured contact resistance at each electrode and used saline water to reduce resistance to under 5000 ohm if necessary. Reducing contact resistance enables higher injected currents, which increases data accuracy and improves signal-to-noise ratios, particularly at depth. We established location and elevation control at a subset of electrodes through post-processing of data collected with a Juniper Systems Geode GPS receiver and observed an average vertical accuracy of 50 cm. Relative elevation differences along each ERT transect were measured with a Leica DISTO.

We processed and terrain-corrected the ERT survey results for elevation changes using EarthImager 2D software [34]. This processing includes an inspection of raw data for quality assurance and an iterative process of 2D inversion modeling and noisy data filtering. We used a smooth model inversion which requires the smooth model complexity to be equal to or less than the true model complexity [35]. During this iterative inversion modeling process, noisy data points are removed until the root mean square error (RMS) is below 10%; each survey required five to eight iterations resulting in RMS below 5% and the retention of between 81% and 96% of data points. The normalized L2-norm parameter, another measurement of data misfit that indicates model convergence at values less than or equal to one, ranged from 0.9 to 1.73 [34]. Regardless of such favorable data misfit, nonuniqueness is an inherent limitation of the ERT method and challenges the reliability of inversion models and subsurface interpretations [36]. This nonuniqueness is illustrated by the many different inversion models that could explain any given set of ERT survey data and the many different model interpretations that could explain any given inversion model. As is common practice, the study reduced nonuniqueness and increased confidence in models and interpretations by carefully selecting inversion parameters and incorporating observed data (i.e. water-level data, borehole lithology, and local field geology).

We acquired lithologic drilling logs for six monitoring wells in the study area: one (37W) is co-located with ERT measurements at the easternmost transect (BRCA1), four are within approximately 500 m of BRCA1, and another (66W) is roughly 2400 m west of the westernmost transect (Figure 1). The five wells clustered near the east and central transects are completed in sandstone, presumably the Straight Cliffs Formation. The wells range in depth from 44.2 to 59.4 m; the alluvium–bedrock interface is found at depths of 9.1 to 21.3 m below ground surface (bgs) (Table 1,1). Depth to bedrock generally increases moving northward away from the fault zone as the alluvium thickens. At the well directly overlapping with geophysical measurements (37W), bedrock is encountered at 9.1 m bgs. The well along fault strike west of the geophysical measurements (66W) is 15.2 m deep and completed in alluvium.

Water-level measurements in the study area ranged from 3.8 to 5.5 m bgs in May 2022 (excluding wells 328W and 329W), 3.5 to 7.3 m bgs in June 2022, and 3.8 to 7.5 m bgs in September 2022. Water levels in the study area generally decrease moving away from the Rubys Inn fault toward the center of Emery Valley; this is supported by Wallace et al. [2] who created potentiometric surfaces for the entirety of Emery Valley. At the well directly overlapping with geophysical measurements (37W), groundwater was encountered at 3.5–3.8 m bgs. Historical records for these wells indicate that water levels have declined by ~1.1 m since the 1980s [19].

For ease of comparison across seasons, we rescaled modeled resistivity ranges for repeat transects to the scale of the survey with the lowest range of modeled resistivities. For the east transect, we rescaled survey lines BRCA1 and BRCA7 to the scale of BRCA1B (Figure 3). For the central transect, we rescaled survey line BRCA3B to the scale of BRCA3 (Figure 4).

The resistivity models for BRCA1 and BRCA1B show the same structure and differ only in the maximum resistivities measured. The models indicate two main geologic layers were imaged northwest of the Rubys Inn fault zone (hanging wall) and five layers were imaged southeast of the fault zone (footwall) (Figure 3,3). A low-resistivity layer with values ranging from about 20 to 80 ohm-m exists at the surface northwest of the fault. This layer has resistivity values consistent with that of saturated, possibly clay-rich alluvium and extends to a depth of about 10 m bgs. At 10 m depth, the resistivity increases to about 120–600 ohm-m, consistent with sandstone and presumably the Straight Cliffs Formation. This change in resistivity at 10 m depth is also consistent with lithologic data indicating bedrock from well 37W, which is located approximately 50 m down the transect from the northwest end (Figure 3,3). Water-level data from this well indicate that the water table is at 3.5 to 4 m depth; however, the shallow unsaturated zone is difficult to identify in the 10 m electrode spacing transects because this array is incapable of resolving features smaller than 5 m. This feature is more accurately resolved in the 4 m electrode spacing in transect BRCA7 (Figure 3,3). Approaching the fault zone, both the shallow conductive layer and deeper resistive layer appear to dip more steeply to the southeast (Figure 3,3), consistent with mapping in the area [7]. Additionally, surficial deposits at this point in the transect transition from Quaternary alluvium and colluvium (Qac) to a more resistive (100, 600 ohm-m) presumably unsaturated Quaternary colluvium (Qc). Across the fault on the footwall, the data indicate five layers of differing resistivities, which we interpret as alternating layers of the white member and possibly underlying pink member of the Claron Formation. The uppermost and lowermost layers are highly resistive (200, 9000 ohm-m in BRCA1 and 200–3000 ohm-m in BRCA1B) and likely limestone-dominant. Another possible lithology for the uppermost layer is the Tertiary conglomerate at Boat Mesa, which overlies the Claron and outcrops to the east-southeast of the study area. The remaining three layers are poorly to moderately resistive, ranging (from top to bottom) from 15–60 ohm-m to 60–200 ohm-m to 30–80 ohm-m. We interpret this as interbedded mudstone and siltstone within the Claron Formation. Without direct subsurface observations on this side of the fault, it is difficult to conclude whether the lower resistivity layers are saturated or unsaturated and clay rich. Also evident in the transect is the abrupt transition from steeply southeast-dipping layers to relatively flat layers within a short distance from the interpreted fault (~100 m), which is a well-documented feature of the Rubys Inn fault in this area [11]. The Rubys Inn fault zone itself appears to be ~80–100 m northwest of where the east transect crosses the mapped fault. At ~35 m depth, this location has a zone of high-resistivity material ranging from 800 to 20,000 ohm-m. The high resistivity observed here could be evidence of a calcite-cemented fault core and damage zone, or of steeply dipping limestone bedding of the Claron Formation. Vertical to overturned bedding of the Claron Formation is documented to the northeast near the rim of Bryce Canyon National Park [11].

Transect BRCA7 is centered on the fault zone as mapped by Biek et al. [7] and the apparent fault zone as imaged by transect BRCA1. The reduced electrode spacing (4 m) and resulting higher resolution shows similar structures to the 10 m electrode spacing surveys but in greater detail (compare Figure 3,3). At the surface of the northwest end of the transect, the data reveal a shallow (2 and 3 m) upper layer of moderate-resistivity (80, 150 ohm-m) material above a low-resistivity zone; consistent with transects BRCA1 and BRCA1B, we interpret this as unsaturated alluvium overlying saturated alluvium.

The ERT models identify both the groundwater-level and alluvium–bedrock contact in the hanging wall to an excellent degree, yielding refined contacts and geometries, especially BRCA7 (Figure 3,3). Although the land surface gradient and potentiometric surface gradient of Wallace et al. [2] are to the northwest, the contact between what we identify as saturated alluvium and Straight Cliffs Formation dips to the southeast near and northwest of what we interpret as the primary fault plane. The low-resistivity layer (20 and 80 ohm-m) continues horizontally across the fault zone before dipping steeply to the southeast again. Across the fault zone in the footwall, we interpret this low-resistivity layer as a steeply dipping drag fold of clay-rich mudstone or siltstone facies of the Claron Formation, but it is plausible that the low resistivity could also be attributed to flow from what we know is saturated alluvium above the hanging wall. Without well data closer to the fault zone or in the footwall, the location of groundwater and hydraulic gradient are uncertain.

Similar to BRCA1 and BRCA1B, the resistivity models for BRCA3 and BRCA3B generally differ mainly in maximum modeled resistivities. However, these models also exhibit seasonal hydrologic variation, discussed below. At the surface northwest of the mapped fault location is a ~4- to 5-m-thick layer of moderately high-resistivity (100 and 400 ohm-m) material that we interpret as unsaturated alluvium (Figure 4,4). Below this is a ~6- to 9-m-thick layer with lower resistivity (40 and 100 ohm-m) values consistent with saturated alluvium. At the bottom of this layer, at a depth of ~11–13 m, is a thick layer with resistivities of 170–400 ohm-m. This resistivity is consistent with sandstone of the Straight Cliffs Formation as interpreted in transects BRCA1 and BRCA1B. The surface at the southeast end of the transect is a moderate-resistivity layer (75 and 200 ohm-m) that we interpret as ~3 m of unsaturated alluvium. Below this layer are two highly contrasting offset layers: a relatively high-resistivity layer (150 and 400 ohm-m) that is potentially limestone of the white member of the Claron Formation, and a low-resistivity layer (15 and 50 ohm-m) that is potentially saturated alluvium, saturated weathered limestone of the Claron Formation, or mudstone (saturated or unsaturated) of the Claron Formation. Without lithologic control on the footwall side of the fault, we are unsure of the thickness of the alluvium overlying the Claron Formation. A man-made pond is in a swale to the east-southeast of the transect; infiltration and lateral movement of water from this pond could be the source for saturation of the sediment and/or rock in the low-resistivity layer. The center-right of the transect has a northwest-dipping region of highly contrasting resistivities; we interpret this as the primary fault plane. In the May resistivity model, there is a shallow low-resistivity zone coincident with a topographic low that is absent or diminished in the September resistivity model. In September, the low-resistivity layer that we identify as potentially saturated material appears to extend deeper into the subsurface along the fault plane. We suggest that this seasonal resistivity change shows downward groundwater movement through the footwall along the fault. Two other possibilities are that this groundwater movement is occurring on the hanging wall side of the fault, or through a damage zone along the fault plane. Although the fault is concealed at this location, the apparent fault location is ~70 m southeast of the mapped fault location. We see little evidence of bedrock dipping near the fault zone on either side, despite the transect being only ~300 m away from the east transect which exhibits evidence of dipping strata on both sides of the fault in resistivity models.

The ERT inversion models adequately depict the groundwater level and depth to bedrock in the hanging wall based on water-level and lithologic constraints from nearby wells. Conversely, without observed data in the footwall, we are uncertain of the lithology of the near-surface low-resistivity layer (15 and 50 ohm-m) (Figure 4,4). Given the proximity to the Quaternary fluvial channel (Qac on Figure 1,1) and pond to the southeast, the preferred interpretation is that the low-resistivity zone consists of saturated alluvium, although the depth of the underlying sediment–bedrock interface is unclear. Additionally, given the seasonal variation in location and extent of the low-resistivity layer that trends downward along the interpreted fault plane, the ERT inversion models suggest that groundwater may be flowing to the northwest down-dip along the fault. However, another possible interpretation is that the low resistivity along the interpreted fault plane represents clay-rich fault gouge in the damage zone of the footwall, discussed further below. Determining the lithology and hydrogeologic control of low-resistivity material in the fault zone requires further study (e.g. direct observation of subsurface materials from drilling and aquifer testing).

The resistivity model for BRCA5 shows horizontal to subhorizontal layering on each side of the interpreted fault zone (Figure 5,5). North of the fault zone in the hanging wall, we interpret an upper high-resistivity layer (200 and 10,000 ohm-m) as unsaturated colluvium and/or alluvium transitioning to Straight Cliffs Formation. Below this is a low-resistivity layer (5, 30 ohm-m) that is likely Straight Cliffs Formation; the low resistivities could be attributed to groundwater saturation, clay-rich mudstone facies, or both. South of the fault zone in the footwall, we interpret interbedded limestone and mudstone of the pink member of the Claron Formation with alternating high (200 and 800 ohm-m; limestone) and low (5 and 50 ohm-m; mudstone) resistivities. There are no wells on the footwall side of the fault to constrain depth to water, so it is uncertain whether the low-resistivity layers are saturated. The interbedded Claron layers are horizontal in the south of the transect but abruptly transition to steeply south dipping (~45°) in the north-central region of the transect. We interpret this change in dip as a drag fold that truncates at the primary fault plane. Biek et al. [7] mapped nearby Claron strata dipping to the south as much as 80° in the footwall adjacent to the fault. They also mapped Straight Cliffs Formation strata of the hanging wall dipping to the south between 40° and 65°, which is not clearly discerned in the resistivity model. We interpret the resistivity model as showing the surface trace of the fault approximately 25 m north of the location mapped by Biek et al. [7].

Without adequate water-level constraint near the west transect, the ERT model does not confidently locate groundwater depth on either side of the thrust fault at the west transect. The resistivity model identifies high-, moderately high-, and low-resistivity layers on both sides of the fault; the highest contrast occurs in the hanging wall, at what is either a sediment/bedrock interface, interface between facies of the Straight Cliffs Formation, or the water table. Unlike the previously discussed transects, the lower contrast between resistivities along the apparent fault plane makes interpretation more difficult. One possibility is that the lowest resistivity zone (<10 ohm-m) shown in the hanging wall (Figure 5,5) is indicative of clay-rich fault gouge. In this case, we would assume reduced permeability and a reduction of across-fault groundwater flow in this region. However, without hydrogeologic control, we cannot reduce uncertainty in this interpretation.

Inversion models for the ERT surveys produce clear depictions of the subsurface for all transects. The high resolution in these data makes features like alternating beds of differing lithology and resistivity, changes in the dip of stratigraphy, or changes in groundwater occurrence readily apparent. For example, both 10-m and 4-m surveys of the east transect depict the transition from horizontal to gently dipping beds in the interpreted footwall drag fold. Deformation in the footwall of the Rubys Inn fault is well documented in Hillsdale Canyon, approximately 15 km to the west of the study area. Cleveland et al. [17] observed deformation bands, cataclastic flexural slip, and plastically deformed petrified wood in the flexural slip zone all within the Wahweap Formation. These processes often reduce permeability [4]. Although there are no sufficient outcrops to observe whether these same processes are apparent in the Straight Cliffs Formation in the study area, the two formations have a similar lithology and formed in a similar depositional environment. Deformation in the fault zone is also well documented at the eastern end of the fault, along Utah Highway 12 in the Tropic Amphitheater (Figure 1,1). Here, Lundin [11] observed minor thrust faults, conjugate fracture sets, and fault breccia in the hanging wall and footwall, both consisting of the pink member of the Claron Formation. He also observed a 10-cm-thick fault gouge zone. We observed this zone of fault gouge, as well as extensive fault breccia and fractures extending into both the hanging wall and footwall (Figures 6,6 and 7,7). While fracturing and brecciation can enhance fault-zone permeability, cataclastic brecciation and fault gouge can reduce permeability [4]. It is important to note that the scale of these features may be below the resolution of the ERT

surveys. Future drilling and geologic sampling can make up for such resolution limitations. Drilling observations will then help reduce uncertainty in ERT-based interpretations away from drilling sites.

Groundwater resources in the Paunsaugunt Plateau and Emery Valley, Garfield County, Utah, support domestic, municipal, and ecological uses in Bryce Canyon National Park and adjacent communities, including Bryce Canyon City. Emery Valley is subject to potential increased water demand from tourism-related development. Such development exacerbates possible climate-related threats to groundwater resources and groundwater-dependent ecosystems in and around Bryce Canyon National Park. The goal of this study was to improve the understanding of the hydrogeologic role of the Rubys Inn fault and whether this fault offers natural protection to nearby springs and seeps from groundwater pumping in Emery Valley.

To identify the presence of groundwater, detect seasonal groundwater-level changes, and characterize the fault zone, we conducted three ERT surveys in May and September 2022. We produced 2D ERT inversion models for each survey line. The Rubys Inn fault is concealed at the surface over a significant portion of the study area. The results of the ERT surveys show that the latest mapped fault location is as much as 70–100 m off from the actual location perpendicular to strike.

Lithologic logs and water-level data from nearby wells support the interpretations of groundwater levels and stratigraphy on the hanging-wall side of the fault. Resistivity models identify saturated and unsaturated alluvium and colluvium overlying Straight Cliffs Formation sandstone and mudstone in the hanging wall and Claron Formation limestone, mudstone, and siltstone in the footwall. Structural features like the primary fault plane and a footwall drag fold are also interpreted in the inversion models.

The ERT results suggest minor hydrologic differences attributed to seasonality of precipitation and groundwater recharge. Depth to groundwater is well constrained by well data in the hanging wall of the east and central transects. Though we identify conductive layers near the surface of the footwall in each transect, without nearby well data, we are uncertain if these represent groundwater or differences in lithology.

This study’s ERT models illustrate the complexity and variability of the Rubys Inn fault within a relatively short distance along strike. Based on the variable contrasts in resistivities and likely lithologic properties observed in this study, as well as diverse observations from previous Rubys Inn fault studies, we can infer that the fault likely has variable barrier/conduit characteristics with respect to groundwater flow. Further investigation into the hydrogeologic role of the Rubys Inn fault is needed to verify this interpretation, including collecting geologic samples via drilling one or more boreholes through the fault plane and/or into the footwall, as well as installing monitoring wells to conduct aquifer tests, collect groundwater samples, and monitor groundwater levels adjacent to the fault zone. ERT results will be critical for accurately siting and appropriately designing such wells.

Studies of fault-zone hydrogeology often use ERT to supplement extensive borehole or other geologic and hydrologic data [30-32]. We show that ERT is a cost-effective and noninvasive tool for detecting the precise surface location and delineating subsurface fault geometry in otherwise data-poor areas. The noninvasive nature of ERT is especially critical in areas that have sensitive ecological or archaeological resources, for example, national parks.

Data associated with this research are available and can be obtained by contacting the corresponding author.

The authors declare that there is no conflict of interest regarding the publication of this paper

This work was funded by the Utah Geological Survey (UGS) and the National Park Service (NPS).

We thank Nicole O’Shea, Jeffrey Hughes, Steve Rice, Tyra Olstad, Will Hurlbut, Skadi Kobe, and Christian Hardwick for valuable assistance with geophysical data collection, processing, and study design. Tyler Knudsen contributed insightful field observations of the study area. We thank Ruby’s Inn for access to their land and monitoring wells. Lucy Jordan and others provided thorough reviews of this manuscript.