Next Article in Journal
Heavy Metal Content Characteristics and Pollution Source Analysis of Shallow Groundwater in Tengzhou Coal Mining Area
Next Article in Special Issue
An Integrated Approach for Saturation Modeling Using Hydraulic Flow Units: Examples from the Upper Messinian Reservoir
Previous Article in Journal
Study and Application of Estimation Method for Environmental Background Pollution Fluxes into the River in Source Areas
Previous Article in Special Issue
Coupled Geophysical and Hydrogeochemical Characterization of a Coastal Aquifer as Tool for a More Efficient Management (Torredembarra, Spain)
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Water
Volume 15
Issue 23
10.3390/w15234090
water-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Identification of Breaches in a Regional Confining Unit Using Electrical Resistivity Methods in Southwestern Tennessee, USA

by
Md Rizwanul Hasan
1,
Daniel Larsen
1,*,
Scott Schoefernacker
2 and
Brian Waldron
3
1
Department of Earth Sciences, The University of Memphis, Johnson Hall, 488 Patterson St., Memphis, TN 38152, USA
2
Protect Our Aquifer, 1910 Madison Ave. #130, Memphis, TN 38104, USA
3
Center for Applied Earth Science and Engineering Research, The University of Memphis, 1000 Wilder Tower, 3675 Alumni Ave., Memphis, TN 38152, USA
*
Author to whom correspondence should be addressed.
Water 2023, 15(23), 4090; https://doi.org/10.3390/w15234090
Submission received: 15 October 2023 / Revised: 17 November 2023 / Accepted: 23 November 2023 / Published: 25 November 2023
(This article belongs to the Special Issue Application of Geophysical Methods for Hydrogeology)
Browse Figures
Versions Notes

Abstract

:
Electrical resistivity and borehole data are applied to delineate lithostratigraphic boundaries and image the geometry of confining-unit breaches in Eocene coastal-plain deposits to evaluate inter-aquifer exchange pathways. Eight dipole–dipole array surveys were carried out, and apparent resistivity was inverted to examine the lateral continuity of lithologic units in different water-saturation and geomorphic settings. In addition, sensitivity analysis of inverted resistivity profiles to electrode spacing was performed. Resistivity profiles from Shelby Farms (SF) highlight the effect of varied electrode spacing (2.5, 5, and 10 m), showing an apparent ~0.63 to 0.75 depth shift in resistivity-layer boundaries when spacing is halved, with the 10 m spacing closely matching borehole stratigraphy. Grays Creek and Presidents Island profiles show clay-rich Eocene Cook Mountain Formation (CMF), with resistivity ranging from 10 to 70 Ω-m, overlying the Eocene Memphis Sand—a prolific water-supply aquifer. Resistivity profiles of SF and Audubon Park reveal sandy Cockfield Formation (CFF) paleochannels inset within and through the CMF, providing hydrogeologic connection between aquifers, and clarifying the sedimentary origin of confining-unit breaches in the region. The results underscore the efficacy of the electrical resistivity method in identifying sand-rich paleochannel discontinuities in a low-resistivity regional confining unit, which may be a common origin of breaches in coastal-plain confining units.

Graphical Abstract

1. Introduction

Non-invasive geophysical methods such as electrical resistivity (ER) are emerging as cost-effective tools for geologic and hydrogeologic studies [1], especially to characterize and delineate subsurface hydrogeologic and geologic structures to improve groundwater management [2,3]. ER has been used to identify the aquifers and aquitards in different geological settings [4,5,6], which is crucial for groundwater management. Defining aquifers and confining-unit boundaries using geophysical methods helps to create better hydrogeologic models for improving groundwater budgets and identifying groundwater contamination flow paths [7,8,9]. Characterizing and identifying a confining unit, especially the layer above a semi-confined aquifer, is critical to prevent contaminants from flowing into an underlying aquifer. The presence of confining-unit breaches (thinning or absence of confining clay layer) raises the potential for groundwater contamination.
The dynamic nature of the study area’s coastal-plain depositional setting has led to the development of a complex stratigraphic framework. Over time, sedimentation processes, erosion, and the influence of relative-sea-level change have shaped subsurface geology, resulting in a heterogeneous distribution of geological formations. Previous studies provide evidence for breaches in the upper Claiborne confining unit (UCCU), which includes the clay-rich Cook Mountain Formation and heterolithic Cockfield Formation, overlying the Memphis aquifer, mainly comprising the Memphis Sand [10,11,12,13,14,15]. The Memphis aquifer is a regionally important water-supply aquifer in Memphis, Shelby County, Tennessee, and most of the northern Mississippi embayment region [16,17]. Recent studies suggest that the breaches may result from the paleo-channel incision during Eocene sea-level fall and Quaternary incision by western Tennessee tributaries to the Mississippi River [18,19].
ER surveys of the subsurface primarily depend on lithology and pore-fluid conditions such as salinity and water saturation. The resistivity of the soil decreases with increasing soil water saturation and pore-fluid salinity. However, the salinity of the pore-fluid affects soil resistivity more than the saturation [20,21,22,23]. Previous studies have highlighted the significance of considering factors such as soil saturation, pore-fluid salinity, and landscape characteristics when analyzing ER data [24,25,26,27,28,29]. The lateral and depth resolution of the ER survey data depends on the chosen array type, length, and electrode spacing. The array length is vital for depth of investigation (DOI) and depends on the electrode spacing. Understanding the suitable electrode spacing can help interpret layer boundaries and properties accurately. The effect of different electrode configurations on subsurface investigation and resolution is discussed in Section 2.2.1 and Section 2.2.3.
Numerous studies have employed ER to identify fault zones and groundwater contamination pathways [7,30,31]. Whereas previous research has identified breaches in the UCCU using diverse data sources [11,14,32,33,34], their locations and extents remain poorly constrained. Notably, none have defined the subsurface geometry of breaches or clarified their paleo-channel origin. This study addresses knowledge gaps in understanding the geometry and origins of breaches in the Mississippi embayment region and coastal-plain settings, in general, utilizing ER with borehole data to visualize and conceptualize near-surface hydrogeology and confining-unit breaches. The research employs ER surveys to not only delineate lithostratigraphic units and assess confining-layer continuity but also to confirm breach presence and explore its potential causes. This innovative approach goes beyond traditional applications of ER, such as hydrostratigraphic unit delineation or identifying fault zones and contamination pathways, showcasing the development of an interpretation methodology for breaches in aquitards. The identification of breaches and assessment of their underlying causes underscore the importance of ER in shallow sub-surface hydrogeologic investigation.

2. Materials and Methods

2.1. Geologic and Hydrogeologic Settings

The Mississippi embayment spans eight south-central United States states and consists of unconsolidated aquifers and aquitards [17,35]. Shelby County, Tennessee, situated within the embayment, relies entirely on groundwater for public supply, withdrawing 696,000 m3/day in 2015 [36]. The geologic setting of the study area comprises alluvial floodplains of modern streams underlain by Pleistocene to Holocene sand and gravel with overlying silty alluvium [14,17] (Figure 1). Uplands are prevalent in the remainder of the county, with Pliocene and Pleistocene fluvial-terrace deposits overlain by Pleistocene loess [10,37].
Shelby County has three primary freshwater aquifers: the shallow, Memphis, and Fort Pillow. The shallow aquifer, with a thickness ranging from 0 to 30 m, comprises alluvial and fluvial deposits extending over the entire county [10,16,38,39]. This aquifer encompasses the Mississippi River Valley Alluvial (MRVA) aquifer on the west side of the bluff line [40]. In the eastern county, the shallow aquifer aligns with the unconfined region of the Memphis aquifer, serving as a crucial recharge zone [11].
The UCCU, a regional confining unit for the Memphis aquifer, lies below the shallow aquifer and comprises the Eocene Cook Mountain and Cockfield Formations (older to younger). The thickness of the UCCU ranges from 0 to 61 m. The Cockfield Formation includes sand, silt, clay, and lignite, commonly in one or more fining-upward sequences of sand grading up to silt, clay, and lignite [10,41]. The Cook Mountain Formation is mostly laminated with thinly bedded silty clay and very-fine-grained sand [14,42]. The UCCU overlies and provides confinement to the Memphis aquifer but is thin or absent in the southern and eastern Shelby County [11,15]. The Memphis aquifer is 152–275 m thick and composed mostly of sand with clay and minor lignite. The Flour Island confining unit, underlies the Memphis aquifer, separating the Memphis and Fort Pillow aquifers, and composed of clay, silt, sand, and lignite.
Water 15 04090 g001
Figure 1. (a) Location of the Mississippi embayment. (b) The map shows the study area in Shelby County and survey locations (red squares). The cross-section line A–A’ extends northwest in Arkansas and southeast in Mississippi [11]. (c) Cross-section of Mississippi embayment stratigraphy along the cross-section line A–A’. The black dot indicates the projected intersection of the cross-section line and the Mississippi River in Figure 1b. (Ref. [43] modified Figure 1c from [44]; current revision shows the direction, study areas on (b), and the intersection of the cross-section line on (c)).
Figure 1. (a) Location of the Mississippi embayment. (b) The map shows the study area in Shelby County and survey locations (red squares). The cross-section line A–A’ extends northwest in Arkansas and southeast in Mississippi [11]. (c) Cross-section of Mississippi embayment stratigraphy along the cross-section line A–A’. The black dot indicates the projected intersection of the cross-section line and the Mississippi River in Figure 1b. (Ref. [43] modified Figure 1c from [44]; current revision shows the direction, study areas on (b), and the intersection of the cross-section line on (c)).
Water 15 04090 g001
This research focuses primarily on the UCCU, which locally includes sand-rich breaches that provide little or no confinement to the Memphis aquifer [11,12]. Although most of the recharge to the Memphis aquifer occurs in the region of the subcrop east of Shelby County [16], modeling of focused recharge through breaches contributes as much as 50 percent of the water withdrawn from the Memphis aquifer [13,44], some of which has associated water-quality impacts [11,13].

2.2. Methodology

The ER data collected for this study were from five different locations in Shelby County (Figure 2). Four surveys were conducted at Shelby Farms (SF) and Gray’s Creek (GC) to evaluate the impact of electrode spacing in stratigraphic delineation. Data collected from SF and GC examined if 10 m electrode spacing best corresponded to the interpretation of the top of the layers. Data from Audubon Park (AP) and President’s Island (PI) further investigated suspected breaches suggested by [11]. In addition, two boreholes were drilled at PI to provide control and verify the interpretation of the inverted resistivity data. Geologic and geophysical logs from nearby boreholes are used to constrain the accuracy of the interpretations.

2.2.1. Subsurface Electrical Resistivity

Soil properties, such as porosity, permeability, rock texture and type, liquid chemicals, and saturation, as well as clay content of the subsurface layer, affect the electrical resistivity values (ERV) [26,27]. Among these parameters, the clay content impacts subsurface conductivity most [27,28,29] in sedimentary deposits. A resistivity survey applies electrical current to the subsurface through current electrodes, measuring the resulting potential difference at the surface through potential electrodes. Current flows radially from a single-point source in a homogeneous, isotropic subsurface, developing equipotential surfaces perpendicular to the current flow [45].
Electrical resistivity has developed as a powerful geophysical method for investigating sub-surface stratigraphy in fluvial depositional environments. Fluvial systems, characterized by the dynamic interchange of sediment transport, deposition, and erosion, leave discrete signatures in the subsurface resistivity distribution. Ref. [46] studied the potential of geoelectrical methods for characterizing shallow sediments in riverbeds. Combining sensitivity analysis and measurement configurations, ER maximizes shallow riverbed hydrostratigraphy mapping effectiveness. In recent years, research has shown the utility of ER for stratigraphic properties studies [2,8,30].

2.2.2. Data Collection

ER data were collected using a SuperSting R8 Wi-Fi instrument with 56 channels. Eight surveys were conducted, each serving a specific purpose in the investigation. Five surveys were carried out at SF and GC to examine the influence of different electrode spacing on delineating stratigraphic boundaries. The electrode spacing in these surveys was adjusted to 2.5, 5, and 10 m at SF and 7 and 10 m at GC, resulting in survey lengths varying from 137.5 to 540 m. Both of these locations are in agricultural fields. However, the SF field has numerous monitoring wells and pipelines along the margins of the field.
Three more surveys were conducted to investigate the suspected breach locations identified in previous studies. These surveys utilized an electrode spacing of 10 m, which is shown to be the spacing that produces profiles that best conform to borehole-based stratigraphic boundaries and lithology. Two surveys at PI had survey lengths of 540 m, whereas a roll-along study was performed at AP, covering a distance of 840 m. PI is an agricultural field with a buried petroleum pipeline running between the two survey lines (~100 m from the ends of each line). AP is a golf course with <10 m relief, buried irrigation lines, asphalt and concrete cart paths, modified land surface for greens and fairways, and possibly other shallowly buried infrastructure.
Data points collected along survey lines at PI and SF are 762 and 1524 for AP and GC, respectively. Data collected at GC contain more data points than the other survey lines due to extended data coverage, which allows overlapping data collection to improve the data inversion process. The extended data coverage was not used for the other survey lines due to minimal improvement in the inversion process. Data were collected at different times during 2021 through 2022, depending on the degree of soil saturation and access. The water table varied for different locations but remained consistent for a specific area. The number of survey lines does not necessarily represent the whole research area but comprises a sample of geological settings where breaches were identified using previous studies [10,11,33,34] or suspected from recent borehole data. The number of survey lines is also limited due to land-access permission and lack of accessibility due to tree, power, and supply lines.

2.2.3. Array Selection and Electrode Spacing

Array configuration in an ER survey controls the amount of data collected in the field and influences the interpretation of the inverted resistivity data. The dipole–dipole array [47] was chosen for this study as it is sensitive to lateral variations in the subsurface [48]. Before selecting the dipole–dipole array for this study, preliminary surveys were carried out using different array configurations, including the Wenner and Schlumberger arrays. These initial array choices, however, showed limitations in terms of lateral resolution, particularly concerning the targeted objective of distinguishing lateral variations within a clay layer. The study by [7] also confirms the suitability of using the dipole–dipole array in a subsurface-lateral-variation study. The data acquisition using these arrays yielded comparatively fewer data points when compared with the dipole–dipole arrays. The reduced data quantity within the preliminary survey arrays requires increased confidence in interpolation techniques, thereby introducing additional assumptions into the data analysis process. Thus, the dipole–dipole array is more suitable for identifying vertical contacts between high and low-resistive materials within a depth range in the subsurface. In the present study, confining-unit breaches have high-resistive material (sand or gravel) juxtaposed with low-resistive material like clay.
Understanding electrode spacing is essential for DOI. The DOI is approximately 20% of the array length [45,46]. We also compared the obtained data to verify the DOI, which confirms the above statement (Appendix C). As previous studies have lacked investigation of the optimum electrode spacing in delineating stratigraphic boundaries, this study investigated the 2.5 m, 5 m, 7 m, and 10 m electrode spacing coupled with a dipole–dipole array to find the best balance of depth resolution and survey length for near-surface sediments in the northern Mississippi embayment.

2.2.4. Data Processing

Apparent resistivity data collected in the field were analyzed using AGI EarthImager 2D, V. 2.1.5 – a proprietary software. The contact resistance (CR) and noise percentage were checked for bad data and removed if the CR was above 4000 ohms or the noise percentage was greater than 20%. We chose a smooth inversion model for the initial setting, where data were removed if the maximum error was greater than 5%. Terrain files containing elevation data were applied for terrain correction for the surveys conducted in Audubon Park and President’s Island, except Shelby Farms and Gray’s Creek. Terrain files were not used for Shelby Farms and Grays Creek due to the focus on DOI and electrode-spacing analysis.
Furthermore, terrain variations are <1 m at this location. EarthImager 2D automatically chooses the finite element method for forward modeling when a terrain correction is completed. The default equation solver for the forward is Cholesky Decomposition, which the software manufacturer sets. We selected a maximum of ten iterations and a maximum RMSE of 5% as termination criteria for resistivity inversion settings. The inversion is continued until the RMSE is reduced below 5%. Figure 3 shows the data-processing workflow for this study. The resistivity scale for this study was set to the same scale (10–3500 Ω-m) for a consistent interpretation of the stratigraphic layers.
A zone of no data was present from the distance of 350 to 510 m of the roll-along survey at AP due to triangulation of the overlapped data points below the depth of 55 m (Appendix B). The absence of data points below 55 m of depth and between the distances mentioned above was interpolated during the inversion process.

2.2.5. Data Interpretation

The resistivity range for glacial sediments suggested by [49] was used to interpret the inverted resistivity sections. The ranges for this study are also chosen by comparing the resistivity values collected from the field with nearby borehole data to ensure accuracy in interpretations. The presence of inorganic silt increases the resistivity value of clay [48]. Thus, in this study, ERV less than 50 Ω-m represents clay; 50–100 Ω-m is silty clay with increasing silt and fine sand percentage toward higher ERV; and greater than 100 Ω-m defines sand and gravel, with increasing ERV reflecting decreased silt and clay and increased gravel. The water content and salinity decrease the resistivity of a layer [50,51]. For a 1% increase in water content in silty sand, the ERV decreases by 5% when the water content is ≤35% (calculated from the graph of four electrode methods for silty sand presented in [52]). The changes in the ERV are negligible as water content reaches an equilibrium state [52]. The ERV decreases rapidly when specific conductance (SC) is ≥0.02 mS/cm [51]. At SF, SC ranges from 0.065 to 1.33 mS/cm [53], whereas the SC ranges from 0.045 to 0.140 mS/cm near GC [54] at 25 °C for the alluvial and the Memphis aquifer. Near the PI area, the Memphis aquifer SC ranges from 0.150 to 0.393 mS/cm [14]. At the Sheahan well field near AP, SC ranges from 0.102 to 0.151 mS/cm [32] in the Memphis aquifer. The SC of the MRVA aquifer near the PI area is higher than that of the Memphis aquifer and ranges from 0.55 to 0.89 mS/cm. Hence, the ERV in the MRVA aquifer will be lower than in the Memphis aquifer given that all factors (clay fraction, porosity, salinity) are similar for both.

3. Results

3.1. Shelby Farms

The subsurface geology at the SF site is well constrained by geologic borehole data [12,33,34], shallow seismic surveys [18], and previous ER surveys [7], which mainly focused on the impact of specific conductance from landfill leachate. Thus, SF provides a suitable location for assessing the effect of ER electrode spacing on profile correspondence to stratigraphic units designated using borehole and geophysical data. Analysis of ER profiles based on electrode spacings of 2.5, 5, and 10 m are compared to information obtained from geophysical well logs and geologic descriptions from boreholes Q-151 and Q-125, and the latter of which is ~5 m from the survey line.
The top of the high ERV zone is estimated at approximately 16 m, based on the depth of sand-rich deposits in borehole Q-125, inferred from the contrast in ERV shown in Figure 4a at an electrode spacing of 10 m. The inverted resistivity profile (IRP) generated with a 5 m electrode spacing (Figure 4b) exhibits a high ERV boundary at a depth of 12 m—4 m higher than that of the 10 m electrode spacing. The IRP conducted with a 2.5 m electrode spacing (Figure 4c) shows the top of the high ERV boundary at a depth of 7.5 m. Based on these observations, when the electrode spacing is reduced by half, the depth to the top sand-rich zone decreases by a factor of ~0.75–0.63. Q-125 contains a metal protective casing and is ~5 m from the survey lines. The metal casing can act as a conductor, allowing electrical current to bypass the subsurface and travel directly through the casing [55]. This phenomenon can introduce biases or distortions in the survey data near the well, leading to overestimating subsurface resistivity. Although this effect may be significant for the 2.5 m electrode spacing, it is less likely to be a problem for the 5 m electrode spacing. The apparent decrease in the depth of the high ERV zone is not clearly understood. However, it may be related to the processing or inherent characteristics of the ER technique. Comparison to the borehole-based lithology indicates that the 10 m electrode spacing best represents lithological changes, especially those below 5 m depth.
The profiles from the shorter electrode spacings (2.5 and 5 m) have value because they resolve the upper 10–15 m of subsurface in greater detail than the longer spacing. For example, Figure 4c explicitly shows a low-resistivity zone (<50 Ω-m) within the top 4 m that corresponds well to the upper alluvium and an underlying moderate to high ER layer (ranging from 80 to 200 Ω-m) that fits well to the sand and gravel alluvium. Whereas in Figure 4a, the upper ~10 m are an intermediate ER layer ranging from 80 Ω-m to 200 Ω-m with little discernable structure. Furthermore, the intermediate ER layer is underlain by a high ERV zone, 100 Ω-m to greater than 3500 Ω-m, with noticeable lower resistivity discontinuities (Figure 4a,b). This interval is not apparent in Figure 4c, perhaps due to the distortion caused by the steel casing or poor depth resolution. The size and geometry of the discontinuities are consistent with paleo-channel features identified from shallow-shear-wave seismic analysis along approximately the same line [18]. The revised stratigraphic interpretation of the SF by [19] indicates that the paleo-channel features are in the Eocene Cockfield Formation, based on stratigraphic position and texture.
The high ERV layer (ranging from 100 to 200 Ω-m) observed at depths between 50 and 80 m in Figure 4a is interpreted as the Memphis Sand, which aligns with [56] and recent stratigraphic revisions by [19]. The low ER layer, below 50 Ω-m, observed below 80–90 m (Figure 4a) corresponds to a silty clay layer locally observed in the middle of the Memphis Sand, which may be equivalent to the Zilpha clay defined in Mississippi [17]. The low-resistivity layer is not visible in Figure 4b,c due to the shallower DOI.

3.2. Grays Creek

Electrical resistivity surveys were conducted at the GC site to investigate the resistivity ranges for the Cook Mountain Formation and Memphis Sand and a potential breach inferred from borehole data. Two orthogonal surveys, a long line with 10 m electrode spacing and a shorter one with 7 m electrode spacing, were used to assess further the influence of electrode spacing on the depth of strong resistivity contrasts (the boundary between the top high-resistivity zone and the low-resistivity zone below it).
In the IRPs from GC (Figure 5a,b), the alluvium is a low-to-moderate ERV layer ranging from 15 to 80 Ω-m and is observed in the top 12 m. Similar to the results at SF, the sand-rich lower alluvium is distinct in the IRP with shorter spacing. The low ERV layer (<50 Ω-m) extending from a depth of ~12 to ~32 m is the Cook Mountain Formation. The high ERV interval, ranging from 100 to 300 Ω-m, was observed in both IRPs below ~32 m and represents the Memphis Sand.
Figure 5a,b provides insights into the average depths of the layer boundaries between lithologic units. The average depth for the boundary between the Cook Mountain Formation and Memphis Sand is estimated at ~32 m for the 10 m spacing and ~32.5 m for the 7 m spacing survey, suggesting little or no apparent depth shift. The boundary depth between the Cook Mountain Formation and Memphis Sand, determined by the well log (TN157_000438) ~200 m north of GC line 1, is 32.5 m mbgs (meter below ground surface). Similarly, the depth of the base of the alluvium in both IRPs is ~8 m, which is similar to or a little deeper than that observed in the nearby borehole. The absence of a distinct shift may be related to very high resistivities or the absence of steel casing adjacent to the survey lines.
The inferred boundary between the alluvium and Cook Mountain Formation deepens to the north, which is away from the direction of the stream. The deeper contact is interpreted to reflect alluvium underlain by the Cockfield Formation inset into the Cook Mountain Formation, which is evident in the adjacent well log. Although there is no evidence of the presence of a paleochannel, the inferred shape and slope of the Cockfield—Cook Mountain formation contact to the north towards the well is believed to represent a paleo-channel just north of line 1 at GC. Land inaccessibility limited further assessment of the inferred paleo-channel and potential breach.

3.3. President’s Island

Both IRPs discussed at PI are oriented west to east on President’s Island (Figure 6), a former Mississippi River island now connected to the east bank of the Mississippi River. The primary goal of conducting these two profiles was to investigate a previously proposed breach beneath most of PI [11], recent evidence for which is suggested locally from inverted airborne electromagnetic (AEM) data [29]. Because no lithologic control from borehole data was available at the site, cored boreholes were drilled, and gamma logs (Appendix D) were obtained along each line to provide detailed geologic information (Figure 5).
Line A is positioned ~730 m east of the river. A discontinuous layer with a moderate ERV range of ~50–110 Ω-m is present from 65 to 58 masl (meter above sea level) in Line A (Figure 6a). At 60–20 masl, a high ERV zone (>100 Ω-m) is observed, which connects to the surface at both ends of the survey line. Combining these two intervals forms the MRVA with the low-resistivity upper layer being the fine-grained, silty alluvium. At ~34 to 20 masl, a low ERV layer (<50 Ω-m) underlies the MRVA with ~14 m of relief along the contact. This unit extends to ~23 m masl or more. It is represented by laminated-to-thinly-bedded silty clay and silty, very-fine-grained sand of the Cook Mountain Formation in borehole TN157_003511 (Figure 7). The lower part of the IRP, below −23 masl, in Figure 6a had low-quality data removed from the profile and not shown but likely includes high-resistivity Memphis Sand strata.
Line B is 1.25 km east of the Mississippi River and 255 m east of Line A. The IRP of Line B (Figure 6b) exhibits a moderate ERV zone of the silty alluvium to 120 m towards the east from the western side of the line. Below the intermediate ERV layer and along the remainder of the line to the east, a high ERV zone is observed between ~57 and ~14 masl, representing sand and gravel in the MRVA. A low-to-moderate ERV interval cuts across the high ERV zone from 240 to 170 m along the line. In borehole TN157_003510, this interval includes more silt and clay. The overall shape is consistent with a buried MR channel margin, similar to those inferred at the Tennessee Valley Authority (TVA) sites [14]. Like line A, the base of the high ERV zone has ~19 m of relief, with deepening of the contact to the east. Ref. [14] also noted the deepest level of incision of the MRVA in the underlying Eocene strata, which was deepest along the eastern side of the TVA sites. From ~34 masl to ~−40 masl, a low ERV layer underlies the MRVA that is represented by laminated-to-thinly-bedded silty clay and silty, very-fine-grained sand of the Cook Mountain Formation in borehole TN157_003510. The intermediate-to-high resistivity in the interval of ~26 to ~8 masl at the borehole site may represent finer sand in the upper part of the Cook Mountain Formation in that area. The intermediate-to-high ERV layer below the Cook Mountain Formation in the center of line B is the Memphis Sand, which is better resolved in line B than in line A.
The PI ER lines illustrate the stratigraphy beneath the Mississippi River floodplain and do not support what [11] previously interpreted as a broad breach. The silty alluvium is not represented as continuous based on the ER lines, which contrasts with borehole data collected at PI and data from the TVA sites. However, it is not uncommon to have poor resolution in the upper 5 m of an ER survey with 10 m electrode spacing, similar to that observed at the SF sites (Figure 4). The geometry of the sand and gravel of the lower MRVA at PI is consistent with that inferred at the TVA sites, with deeper incision and thickening of the MRVA to the east. Furthermore, line B illustrates the probable channel geometry of one of the channel complexes defined at the TVA sites. Like the TVA sites, Cook Mountain is below the MRVA, and the Cockfield Formation is absent. Although intermediate-to-high ER is observed locally in line B, the sediments representing these deposits are silty and clay-rich and do not represent well-sorted sand typical of incised paleo-channel deposits, such as that observed at SF (Figure 4). Although high-resistivity intervals are observed in AEM data beneath PI [29], they are in inaccessible areas west of lines A and B. The results of the ER study suggest that low-to-intermediate ERV intervals beneath the PI site do not represent breach sites. The extent of the breach at PI is likely much smaller than that [11] proposed initially.

3.4. Audubon Park

AP is situated in the heart of Memphis near the Sheahan well field, one of eleven Memphis’ Light Gas and Water municipal-well fields. Numerous studies have been conducted using geochemical tracer, hydrologic, and borehole data on the leakage of modern water into the Memphis aquifer near the Sheahan well field [32,57,58]. The water table elevation is 60 masl [59] at the AP site and is depressed relative to the water table elsewhere in the Memphis area, suggesting water leakage from the shallow aquifer to the Memphis aquifer [32]. Two boreholes with geologic and geophysical logs are aligned with the ER line (TN157_000105 and TN157_000112), and a cored borehole is shown (TN157_003082) 0.5 km west of the line (Figure 8).
Within the uppermost 10 m, a layer with a variable ERV range of ~10 to 90 Ω-m is observed, with localized intermediate-to-high ERV zones of ~100 to 130 Ω-m (Figure 8). The low ERV zones likely represent loess with variable saturation; the lowest values represent more saturation. The spacing of low-resistivity maxima is periodic, with most separated by 10 to 30 m of intermediate-to-high resistivity, presumably representing drier zones. Below ~85 masl, a high ERV layer exceeding 300 Ω-m becomes prominent but is segmented by the intermediate-to-high ERV zones observed in the overlying loess. The upper part of the high-resistivity layer corresponds to sand and gravel of the Pliocene Upland Complex, a fluvial-terrace deposit that underlies most upland surfaces in Shelby County [37] or Pleistocene fluvial-terrace paleo-channel-fill deposits beneath the Sheahan well field area [60]. The source of the irregular variation in the high ERV layer is unclear. However, it may represent variable saturation or lithologic changes, especially clay accumulation due to paleosol development in the upper part of the sand and gravel. The base of the fluvial-terrace deposits is interpreted to be in the lower part of the highest ERV zones (>1000 Ω-m) based on geophysical and geologic logs. The lower part of the high ERV zone is the sand-rich Cockfield Formation based on the geologic borehole log from TN157_0003082 and geophysical logs from the other two boreholes. Between 160 and 250 m and 620 and 690 m, the fluvial terrace is underlain by a low-to-intermediate resistivity unit interpreted as the Cook Mountain Formation. The sand-rich Cockfield Formation is inset into the Cook Mountain Formation and from 290 to 570 m inset into the Memphis Sand. These characteristics are consistent with the Cockfield Formation being inset as paleochannel deposits into the underlying Eocene units. In all cases, the paleochannel has a concave-up geometry. The intermediate-to-high ERV interval beneath the Cockfield Formation represents the Memphis Sand based on the cored borehole log from TN157_0003082 and geophysical logs from the other two boreholes. The Memphis Sand in borehole TN157_0003082 has intervals of fine sand and clay and clay ball conglomerate, likely contributing to the intermediate ERV.
The AP line generally follows the lithological variations expected from the stratigraphy determined from cored borehole TN157_0003082 and geophysical logs from other boreholes. However, it illustrates complicated patterns likely due to surface features and irregular saturation. Figure 8 shows numerous surface features along the line on the AP golf course. In addition, an underground sprinkler system and modification to the surface beneath putting greens, which are not shown, may also influence the signals. The distribution of the sprinklers, in particular, may create the periodic patterns of low resistivity observed in the upper 10 m of the line. The extension of the intermediate resistivity zones from the surface into the fluvial-terrace deposits may also have origins in the surface features, but as stated above, lithology and saturation could also play a role. The extensive modification of the surface and greater depth to the water table at the golf course, as opposed to the rather monotonous surface conditions with a shallow water table beneath the agricultural fields, especially at GC and PI, result in complications to the resulting IRPs, which require ample geologic control based on borehole data to constrain interpretation.
The IRP from the AP line supports the absence of a confining unit beneath much of the golf course, proximal to the ER line. It is consistent with borehole geophysical log data [58] and water level and hydrologic tracer studies [13,32,58]. Desaturation of the shallow aquifer and development of an anomalous depression in the water table [11] requires a direct hydrologic pathway from the fluvial-terrace deposits to the Memphis Sand, which is provided by the Cockfield paleo-channel system (Figure 8). Ref. [57] previously proposed that the confining unit was absent in a shallow borehole adjacent to borehole TN157_0003082, showing the influence of oxygenated shallow groundwater infiltrating the Memphis aquifer. Most recently, Ref. [58] used MODFLOW to model modern water through a breach beneath the AP golf course and found a favorable comparison to hydrologic tracer data. The AP ER survey imaged the physical manifestation of at least a part of the system of breaches beneath the Sheahan well field that have led to drainage of the fluvial-terrace aquifer and leakage of modern water into the Sheahan well field.

4. Discussion

4.1. Effect of Electrode Spacing

Analysis of the inverted resistivity data obtained from the ER surveys at SF reveals that electrode spacing can significantly impact the interpretation of hydrostratigraphic features. Notably, different electrode-spacing configurations lead to an apparent upward shift in the ERV layer boundaries representing various subsurface layers. The extent of this impact is more pronounced at SF (0.75–0.63 times) when the electrode spacing is reduced to half of its initial value. It is unclear why the ERV domain shifts boundaries for spacings at SF; however, the sensitivity appears to be greatest at depths < 10 m using the dipole–dipole array, similar to the results of [3]. Steel casing near the survey lines at SF may introduce errors in the depth calculation of the stratigraphic units [55], but this is likely only a local effect. The influence of the steel casing on the electrical resistivity values around the well needs to be further investigated, as well as the potential influence of other infrastructure at SF (e.g., [7]). The depth shift may also be possible with shorter electrode spacing due to a significant decrease in the apparent resistivity values, due to a shallow conductive layer at a low acquisition level [61].
Similar to the previous study [7], this study shows that an electrode spacing of 10 m yields ERV layer values and boundaries most consistent with local stratigraphy. This particular spacing configuration yields ERV layer boundary depths consistent with boundaries of subsurface stratigraphic units based on borehole data. Utilizing a larger electrode spacing yields resistivity measurements that predominantly reflect the deeper stratigraphic units, with commensurate data loss at depths < ~5 m. The results from the SF and AP surveys suggest, however, that surface and shallow subsurface features influence the ERV at depths > 5 m, suggesting that surveys using short (~2.5 m) and long (~10 m) electrode spacings with the dipole–dipole array will increase or decrease the overall layer resistivity.

4.2. Electrical Resistivity Data Analysis

At SF and AP (Figure 4a and Figure 8), the IRPs indicate the absence of the Cook Mountain Formation or substantial clay in the Cockfield Formation between the shallow and Memphis aquifers, indicating confining-unit breaches in these areas. The SF profile reveals paleo-channel features similar to those that [18] identified using shallow seismic methods. The similarity demonstrates the potential compatibility of the two methods to be used in tandem to resolve the subsurface structure and ERV domain that characterize the geometry of paleochannel features, either inset into low-resistivity clay or within predominantly high-resistivity sand. At SF, silty clay of the Cook Mountain Formation is present ~0.5 km west of the ER survey line at the same depth as the paleochannel features [12], suggesting incision and lateral migration of paleo-river systems in the Eocene as an origin of the hydrogeologic breaches. The pattern of paleo-channel overlap suggests channel migration towards the west or southwest through time, eroding the underlying Cook Mountain Formation and placing the sandy paleo-channel fill of the Cockfield Formation in direct contact with the Memphis Sand. However, the study of paleochannel migration is beyond this research’s scope and requires future investigation. The AP profile incision, or partial incision of Cockfield paleochannels through the Cook Mountain Formation, is evident at three locations along the profile. Ref. [58] shows that the paleochannel beneath the AP golf course extends westward into a buried Eocene tributary drainage system. The ER profile indicates little evidence for lateral channel migration; however, perhaps shallow S-wave seismic analysis similar to that conducted at SF by [18] may be useful to resolve such depositional structure.
The ER surveys at GC and PI demonstrate consistency with regional Eocene stratigraphy and similarity of lithologic characteristics across the study area. The lithology of the Cook Mountain Formation is similar across Shelby County [19], which is well supported by the similarity in the ERV range observed. The Cook Mountain Formation is thought to have been deposited across Shelby County as a shallow marine to brackish water prodelta clay [14,19]; thus, the continuity of lithology is expected. Thus, the absence of the Cook Mountain Formation implies removal rather than non-deposition. Paleochannel erosion of coastal-plain prodelta and marine clay units similar to the Cook Mountain Formation is also known from elsewhere in the Gulf and Atlantic coasts [62,63,64], suggesting further investigation may provide a broader mechanistic basis for their origin.

5. Conclusions

This study focuses on the investigation of electrical resistivity (ER) as a tool for investigating shallow stratigraphy (<100 m) in the Mississippi embayment, and defining the characteristics and geometry of breaches through a regional confining unit. ER surveys were conducted in the floodplain (GC, PI, and SF) and upland (AP) settings, some with excellent supporting geologic data and others with limited data. Initial investigations at SF, a well-studied site with numerous boreholes near the survey lines, focused on using electrode spacing (2.5, 5.0, and 10.0 m) to resolve stratigraphic units and compare them to borehole data. Electrode spacing was also investigated at GC, applying 7.0 and 10.0 m spacing. All surveys involved a dipole–dipole array, from which the apparent electrical resistivity values (ERV) were inverted to create inverted resistivity profiles (IRPs) using AGI EarthImager 2D—V. 2.1.5 software.
The electrode-spacing analysis suggests that varying the electrode spacing can result in an apparent vertical shift of ERV boundaries on IRPs. The 10 m electrode spacing yielded ERV boundaries that conform best to borehole-based stratigraphy. At SF, decreasing the electrode spacing by half caused the apparent depth of ERV boundaries to shift to shallower depth by a factor of ~0.63–0.75. A potential local source for the shift may be steel casing from a monitoring well < 5 m from the line. However, whether this is important at greater distances (50 or 100 m) from the steel casing is unclear. At GC, the shift in the depth for the two spacings was insignificant considering the m-scale resolution of the IRPs, suggesting that factors creating the shift may be specific to the SF site or greater deviation in electrode spacing. In both cases, the 10 m electrode spacing yields ERVs and ERV boundaries that best fit with known site stratigraphy. Furthermore, at SF, a channel-shaped ERV boundary noted in the 10 m survey correlated well with a similar structure identified by [18] using a shallow S-wave seismic survey, suggesting the potential to apply both survey tools to resolve structure and lithology of significant features better. Surface modifications, such as those at SF and the AP golf course, may create additional complications to interpreting IRPs and may be fruitful investigations for surveys with short (<5 m) electrode spacing.
The ER surveys at SF and AP identified paleochannel features within the sand-rich Cockfield Formation that have incised partially or entirely through the clay-rich Cook Mountain Formation and suggest that paleochannel incision may be a significant origin for breaches in the regional confining unit. The IRPs from SF show sand-filled paleochannel features within the Cockfield Formation directly overlying the Memphis Sand. The Cook Mountain Formation is present at the same depth in boreholes ~0.5 km away, suggesting that paleochannel incision removed the Cook Mountain Formation and the subsequent sand fill of the Cockfield Formation created the breach. More clearly at AP, the IRP shows sand-rich Cockfield Formation filling the paleo-channel structures incised into and through the Cook Mountain Formation and into the Memphis Sand. In both cases, the sand-rich Cockfield Formation in direct contact with the Memphis Sand creates a hydrogeologic connection between the alluvial- or fluvial-terrace aquifer and the regionally important Memphis aquifer. The paleo-channel origin implies that these hydrogeologic connections may have greater lateral extent, along the paleo-channel length, but such inference requires future investigation.
The GC and PI IRPs show a continuous Cook Mountain Formation with similar characteristics between the Memphis Sand and alluvial deposits, which supports the interpretation of the Cook Mountain as the primary confining unit for the Memphis aquifer. The regional extent and shallow marine origin of the Cook Mountain Formation imply that its absence is due to removal rather than non-deposition. Paleochannel incision through regional marine confining units in coastal-plain settings similar to those in the Eocene of southwestern Tennessee suggests that paleochannel breaches in marine confining units are more common than is currently recognized and may have a common origin.

Author Contributions

Conceptualization, M.R.H., D.L., S.S. and B.W.; methodology, M.R.H., D.L. and S.S.; software, M.R.H.; validation, M.R.H., D.L. and S.S.; formal analysis, M.R.H.; investigation, M.R.H.; resources, D.L. and S.S.; data curation, M.R.H.; writing—original draft preparation, M.R.H.; writing—review and editing, D.L., S.S. and B.W.; visualization, M.R.H. and D.L.; supervision, D.L.; project administration, S.S.; funding acquisition, B.W. All authors have read and agreed to the published version of the manuscript.

Funding

This study was part of the aquitard study project and was funded by Memphis Light, Gas, and Water (MLGW), contract number 12064.

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available due to compliance with the MLGW contract.

Acknowledgments

We thank MLGW for financing the research. We thank the Center for Applied Earth Science and Engineering Research (CAESER) and the Center for Earthquake Research and Information (CERI) for providing continuous support with the equipment and planning needed for this research. Special thanks to Kaushik Sarker, a graduate student at CERI, for his help and support during equipment setup and field data collection. We also thank the graduate students who helped us in the fields during data collection. We thank three anonymous reviewers for their helpful comments that improved the clarity and technical aspects of the manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

Appendix A. Comparison of the Inverted Resistivity Profiles of the Data Collected Using Different Arrays

The inverted resistivity profile of the data collected using the dipole–dipole arrays resolves lateral variation better than the data collected using other arrays.
Water 15 04090 g0a1
Figure A1. The inverted resistivity profile of the data was collected using the Wenner array with 10 m electrode spacing and 28 channels at SF.
Figure A1. The inverted resistivity profile of the data was collected using the Wenner array with 10 m electrode spacing and 28 channels at SF.
Water 15 04090 g0a1
Water 15 04090 g0a2
Figure A2. The inverted resistivity profile of the data was collected using the dipole–dipole array with 10 m electrode spacing and 28 channels at SF.
Figure A2. The inverted resistivity profile of the data was collected using the dipole–dipole array with 10 m electrode spacing and 28 channels at SF.
Water 15 04090 g0a2
Water 15 04090 g0a3
Figure A3. The inverted resistivity profile of the data was collected using the Schlumberger array with 10 m electrode spacing and 28 channels at SF.
Figure A3. The inverted resistivity profile of the data was collected using the Schlumberger array with 10 m electrode spacing and 28 channels at SF.
Water 15 04090 g0a3
Water 15 04090 g0a4
Figure A4. The inverted resistivity profile of the data was collected using the Inverse Schlumberger array with 10 m electrode spacing and 28 channels at SF.
Figure A4. The inverted resistivity profile of the data was collected using the Inverse Schlumberger array with 10 m electrode spacing and 28 channels at SF.
Water 15 04090 g0a4

Appendix B. Apparent Resistivity Data Was Collected Using a Roll-Along Survey at Audubon Park (AP), Showing a Triangulated Zone of No Data

Water 15 04090 g0a5
Figure A5. Apparent resistivity pseudosection of the data collected at AP using roll-along survey.
Figure A5. Apparent resistivity pseudosection of the data collected at AP using roll-along survey.
Water 15 04090 g0a5

Appendix C. Verification of Maximum Depth of Investigation

Table A1. Analysis of the DOI and array length relationship at Shelby Farms.
Table A1. Analysis of the DOI and array length relationship at Shelby Farms.
Array Length, L (m)Depth of the Deepest Data Point, D (m)D/L
270590.22
270520.19
13525.90.19
Table A2. Analysis of DOI and array length relationship at Grays Creek.
Table A2. Analysis of DOI and array length relationship at Grays Creek.
Array Length, L (m)Depth of the Deepest Data Point, D (m)D/L
5501190.22
385780.20
Table A3. Analysis of DOI and array length relationship at President’s Island.
Table A3. Analysis of DOI and array length relationship at President’s Island.
Array Length, L (m)Depth of the Deepest Data Point, D (m)D/L
440880.20
5501110.20
5501050.19

Appendix D. Gamma Logs of Boreholes Drilled at President’s Island

Water 15 04090 g0a6
Figure A6. Gamma logs obtained at Pesident’s Island. (a) Borehole TN157_003510, and (b) Borehole TN157_003511. Red line is gamma log down borehole and gray is up the borehole.
Figure A6. Gamma logs obtained at Pesident’s Island. (a) Borehole TN157_003510, and (b) Borehole TN157_003511. Red line is gamma log down borehole and gray is up the borehole.
Water 15 04090 g0a6

References

  1. Kowalczyk, S.; Maślakowski, M.; Tucholka, P. Determination of the Correlation between the Electrical Resistivity of Non-Cohesive Soils and the Degree of Compaction. J. Appl. Geophys. 2014, 110, 43–50. [Google Scholar] [CrossRef]
  2. Casagrande, M.F.S.; Furlan, L.M.; Moreira, C.A.; Rosa, F.T.G.; Rosolen, V. Non-Invasive Methods in the Identification of Hydrological Ecosystem Services of a Tropical Isolated Wetland (Brazilian Study Case). Environ. Chall. 2021, 5, 100233. [Google Scholar] [CrossRef]
  3. Singh, U.; Sharma, P.K. Non-Invasive Subsurface Groundwater Exploration Techniques BT—Environmental Processes and Management: Tools and Practices for Groundwater; Shukla, P., Singh, P., Singh, R.M., Eds.; Springer International Publishing: Cham, Switzerland, 2023; pp. 1–16. ISBN 978-3-031-20208-7. [Google Scholar]
  4. Fajana, A.O. Groundwater Aquifer Potential Using Electrical Resistivity Method and Porosity Calculation: A Case Study. NRIAG J. Astron. Geophys. 2020, 9, 168–175. [Google Scholar] [CrossRef]
  5. Wahab, S.; Saibi, H.; Mizunaga, H. Groundwater Aquifer Detection Using the Electrical Resistivity Method at Ito Campus, Kyushu University (Fukuoka, Japan). Geosci. Lett. 2021, 8, 15. [Google Scholar] [CrossRef]
  6. Wu, J.; Dai, F.; Liu, P.; Huang, Z.; Meng, L. Application of the Electrical Resistivity Tomography in Groundwater Detection on Loess Plateau. Sci. Rep. 2023, 13, 4821. [Google Scholar] [CrossRef] [PubMed]
  7. Schoefernacker, S. Evaluation and Evolution of a Groundwater Contaminant Plume at the Former Shelby County Landfill, Memphis, Tennessee. Ph.D. Dissertation, The University of Memphis, Memphis, TN, USA, 2018. [Google Scholar]
  8. Hussain, Y.; Uagoda, R.; Borges, W.; Nunes, J.; Hamza, O.; Condori, C.; Aslam, K.; Dou, J.; Cárdenas-Soto, M. The Potential Use of Geophysical Methods to Identify Cavities, Sinkholes and Pathways for Water Infiltration. Water 2020, 12, 2289. [Google Scholar] [CrossRef]
  9. Thomas, J.C.; Spring, M.A.; Gruhn, L.R.; Bristow, E.L. Application of Geophysical Methods to Enhance Aquifer Characterization and Groundwater-Flow. Model. Development, Des. Moines River Alluvial Aquifer, Des. Moines, Iowa, 2022; US Geological Survey: Reston, VA, USA, 2023. [CrossRef]
  10. Graham, D.D.; Parks, W.S. Potential for Leakage among Principal Aquifers in the Memphis Area, Tennessee; Water-Resources Investigations Report 85-4295; US Geological Survey: Reston, VA, USA, 1986. [CrossRef]
  11. Parks, W.S. Hydrogeology and Preliminary Assessment of the Potential for Contamination of the Memphis Aquifer in the Memphis Area, Tennessee; Water-Resources Investigations Report 90-4092; US Geological Survey: Reston, VA, USA, 1990. [CrossRef]
  12. Gentry, R.; McKay, L.; Thonnard, N.; Anderson, J.L.; Larsen, D.; Carmichael, J.K.; Solomon, K. Novel Techniques for Investigating Recharge to the Memphis Aquifer; American Water Works Association: Denver, CO, USA, 2006; p. 97. [Google Scholar]
  13. Larsen, D.; Waldron, B.; Schoefernacker, S.; Gallo, H.; Koban, J.; Bradshaw, E. Application of Environmental Tracers in the Memphis Aquifer and Implication for Sustainability of Groundwater Resources in the Memphis Metropolitan Area, Tennessee. J. Contemp. Water Res. Educ. 2016, 159, 78–104. [Google Scholar] [CrossRef]
  14. Carmichael, J.K.; Kingsbury, J.A.; Larsen, D.; Schoefernacker, S. Preliminary Evaluation of the Hydrogeology and Groundwater Quality of the Mississippi River Valley Alluvial Aquifer and Memphis Aquifer at the Tennessee Valley Authority Allen Power Plants, Memphis, Shelby County, Tennessee; Open-File Report 2018-1097; US Geological Survey: Reston, VA, USA, 2018.
  15. Larsen, D.; Waldron, B.; Schoefernacker, S. Updated Map of Semi-Confined Conditions in the Memphis Aquifer, Shelby County, Tennessee: A Work in Progress. In Proceedings of the Virtual 2022 Tennessee Water Resources Symposium, Virtual, 20–21 April 2022; TN Section AWRA: Nashville, TN, USA, 2022; p. 2C-3. [Google Scholar]
  16. Parks, W.S.; Carmichael, J.K. Geology and Ground-Water Resources of the Memphis Sand in Western Tennessee; Resources Investigations Report 88-4182; US Geological Survey: Reston, VA, USA, 1990. [CrossRef]
  17. Waldron, B.; Larsen, D.; Hannigan, R.; Csontos, R.; Anderson, J.; Dowling, C.; Bouldin, J. Mississippi Embayment Regional Ground Water Study; US Environmental Protection Agency: Washington, DC, USA, 2011; p. 192.
  18. Waldron, B.A.; Harris, J.B.; Larsen, D.; Pell, A. Mapping an Aquitard Breach Using Shear-Wave Seismic Reflection. Hydrogeol. J. 2009, 17, 505–517. [Google Scholar] [CrossRef]
  19. Larsen, D.; Schoefernacker, S.R.; Waldron, B. Stratigraphy of Upper Claiborne Strata in Western Tennessee and Hydrogeologic Implications. In Geological Society of America Abstracts with Programs 159-10; The Geological Society of America: Boulder, CO, USA, 2022; Volume 54. [Google Scholar] [CrossRef]
  20. Hong-Jing, J.; Shun-Qun, L.; Lin, L. The Relationship between the Electrical Resistivity and Saturation of Unsaturated Soil. Electron. J. Geotech. Eng. 2014, 19, 3739–3746. [Google Scholar] [CrossRef]
  21. Lu, Y.; Abuel-Naga, H.; Al Rashid, Q.; Hasan, M.F. Effect of Pore-Water Salinity on the Electrical Resistivity of Partially Saturated Compacted Clay Liners. Adv. Mater. Sci. Eng. 2019, 2019, 7974152. [Google Scholar] [CrossRef]
  22. Fatima Zohra, H.; Laredj, N.; Maliki, M.; Missoum, H.; Bendani, K. Laboratory Evaluation of Soil Geotechnical Properties via Electrical Conductivity Evaluación de Laboratorio de Las Propiedades Geotécnicas Del Suelo Mediante Conductividad Eléctrica. Rev. Fac. De Ing. 2019, 90, 101–112. [Google Scholar] [CrossRef]
  23. Corwin, D.L.; Yemoto, K. Salinity: Electrical Conductivity and Total Dissolved Solids. Soil. Sci. Soc. Am. J. 2020, 84, 1442–1461. [Google Scholar] [CrossRef]
  24. Ward, S.H. The Resistivity and Induced Polarization Methods. In Proceedings of the 1st EEGS Symposium on the Application of Geophysics to Engineering and Environmental Problems, Golden, CO, USA, 28–31 March 1988; Environment and Engineering Geophysical Society: Denver, CO, USA, 1998; pp. 109–250. [Google Scholar] [CrossRef]
  25. Zonge, K.; Wynn, J.; Urquhart, S. Resistivity, Induced Polarization, and Complex Resistivity. In Near-Surface Geophysics; Investigations in Geophysics; Butler, D.K., Ed.; Society of Exploration Geophysicists: Tulsa, OK, USA, 2005; pp. 265–300. ISBN 978-1-56080-130-6. [Google Scholar]
  26. Bedrosian, P.A.; Schamper, C.; Auken, E. A Comparison of Helicopter-Borne Electromagnetic Systems for Hydrogeologic Studies. Geophys. Prospect. 2016, 64, 192–215. [Google Scholar] [CrossRef]
  27. Baldridge, W.S.; Cole, G.L.; Robinson, B.A.; Jiracek, G.R. Application of Time-Domain Airborne Electromagnetic Induction to Hydrogeologic Investigations on the Pajarito Plateau, New Mexico, USA. Geophysics 2007, 72, B31–B45. [Google Scholar] [CrossRef]
  28. Wang, H.; Revil, A. Surface Conduction Model for Fractal Porous Media. Geophys. Res. Lett. 2020, 47, e2020GL087553. [Google Scholar] [CrossRef]
  29. Minsley, B.J.; Rigby, J.R.; James, S.R.; Burton, B.L.; Knierim, K.J.; Pace, M.D.M.; Bedrosian, P.A.; Kress, W.H. Airborne Geophysical Surveys of the Lower Mississippi Valley Demonstrate System-Scale Mapping of Subsurface Architecture. Commun. Earth Environ. 2021, 2, 131. [Google Scholar] [CrossRef]
  30. Nabi, A.; Liu, X.; Gong, Z.; Ali, A. Electrical Resistivity Imaging of Active Faults in Palaeoseismology: Case Studies from Karachi Arc, Southern Kirthar Fold Belt, Pakistan. Nriag J. Astron. Geophys. 2020, 9, 116–128. [Google Scholar] [CrossRef]
  31. Porras, D.; Carrasco, J.; Carrasco, P.; González, P.J. Imaging Extensional Fault Systems Using Deep Electrical Resistivity Tomography: A Case Study of the Baza Fault, Betic Cordillera, Spain. J. Appl. Geophys. 2022, 202, 104673. [Google Scholar] [CrossRef]
  32. Larsen, D.; Gentry, R.; Solomon, D. The Geochemistry and Mixing of Leakage in a Semi-Confined Aquifer at a Municipal Well Field, Memphis, Tennessee, USA. Appl. Geochem. 2003, 18, 1043–1063. [Google Scholar] [CrossRef]
  33. Bradley, M.W. Ground-Water Hydrology and the Effects of Vertical Leakage and Leachate Migration on Ground-Water Quality near the Shelby County Landfill, Memphis, Tennessee; Water-Resources Investigations Report 90-4075; US Geological Survey: Reston, VA, USA, 1991; p. 47. [CrossRef]
  34. Parks, W.S.; Mirecki, J.E. Hydrogeology, Ground-Water Quality, and Potential for Water-Supply Contamination near the Shelby County Landfill in Memphis, Tennessee; U.S. Geological Survey Water-Resources Investigations Report 91-4173; US Geological Survey: Reston, VA, USA, 1992; p. 79. [CrossRef]
  35. Clark, B.R.; Hart, R.M. The Mississippi Embayment Regional Aquifer Study (MERAS): Documentation of a Groundwater-Flow Model Constructed to Assess Water Availability in the Mississippi Embayment; Scientific Investigations Report 2009–5172; US Geological Survey: Reston, VA, USA, 2009; p. 61. [CrossRef]
  36. Dieter, C.A.; Maupin, M.A.; Caldwell, R.R.; Harris, M.A.; Ivahnenko, T.I.; Lovelace, J.K.; Barber, N.L.; Linsey, K.S. Estimated Use of Water in the United States in 2015; Circular 1441; US Geological Survey: Reston, VA, USA, 2018; p. 65. [CrossRef]
  37. Van Arsdale, R.B.; Cox, R.T. The Mississippi’s Curious Origins. Sci. Am. 2007, 296, 76B–82B. [Google Scholar] [CrossRef]
  38. Brahana, J.V.; Broshears, R.E. Hydrogeology and Ground-Water Flow in the Memphis and Fort Pillow Aquifers in the Memphis Area, Tennessee; Water-Resources Investigations Report 89-4131; US Geological Survey: Reston, VA, USA, 2001; p. 56. [CrossRef]
  39. Konduru Narsimha, V.K. Altitudes of Water Levels 2005, and Historic Water Level Change in Surficial and Memphis Aquifer, Memphis, Tennessee. Master’s Thesis, The University of Memphis, Memphis, TN, USA, 2007. [Google Scholar]
  40. Lloyd, O.B.; Lyke, W.L. Ground Water Atlas of the United States: Segment 10, Illinois, Indiana, Kentucky, Ohio, Tennessee; Hydrologic Atlas 730; US Geological Survey: Reston, VA, USA, 1995; p. 30. [CrossRef]
  41. Vanderlip, C.A.; Cox, R.T.; Larsen, D.; Mitchell, J.; Harris, J.B.; Cearley, C.S. Newly Recognized Quaternary Surface Faulting and Folding Peripheral to the New Madrid Seismic Zone, Central United States, and Implications for Restraining Bend Models of Intraplate Seismic Zones. J. Geol. 2021, 129, 77–95. [Google Scholar] [CrossRef]
  42. Martin, R.V.; Van Arsdale, R.B. Stratigraphy and Structure of the Eocene Memphis Sand above the Eastern Margin of the Reelfoot Rift in Tennessee, Mississippi, and Arkansas, USA. GSA Bull. 2017, 129, 970–996. [Google Scholar] [CrossRef]
  43. Bursi, J.B. Recharge Pathways and Mechanisms to the Memphis Aquifer. Master’s Thesis, University of Memphis, Memphis, TN, USA, 2015. [Google Scholar]
  44. Brahana, J.V.; Parks, W.S.; Gaydos, M.W. Quality of Water from Freshwater Aquifers and Principal Well Fields in the Memphis Area, Tennessee; Water-Resources Investigations Report 87-4052; US Geological Survey: Reston, VA, USA, 1987; p. 26. [CrossRef]
  45. Jamaluddin; Umar, E. Identification of Subsurface Layer with Wenner-Schlumberger Arrays Configuration Geoelectrical Method. IOP Conf. Ser. Earth Environ. Sci. 2018, 118, 12006. [Google Scholar] [CrossRef]
  46. Clifford, J.; Binley, A. Geophysical Characterization of Riverbed Hydrostratigraphy Using Electrical Resistance Tomography. Near Surf. Geophys. 2010, 8, 493–501. [Google Scholar] [CrossRef]
  47. Lowry, W.F. Geophysical Techniques. In Field Sampling Procedures Manual; The New Jersey Department of Environmental Protection (NJDEP): Trenton, NJ, USA, 2022; p. 46. [Google Scholar]
  48. Wightman, W.E.; Jalinoos, F.; Sirles, P.; Hanna, K. Application of Geophysical Methods to Highway Related Problems; Tech Report FHWA-IF-04-021; US Federal Highway Administration: Lakewood, CO, USA, 2003; p. 744.
  49. Palacky, G.J. Resistivity Characteristics of Geologic Targets. In Electromagnetic Methods in Applied Geophysics–Theory Volume 1; Investigations in Geophysics, Volume 3; Society of Exploration Geophysicists: Houston, TX, USA, 1988; pp. 52–129. ISBN 978-0-931830-51-8. [Google Scholar]
  50. Samouëlian, A.; Cousin, I.; Tabbagh, A.; Bruand, A.; Richard, G. Electrical Resistivity Survey in Soil Science: A Review. Soil. Tillage Res. 2005, 83, 173–193. [Google Scholar] [CrossRef]
  51. Friedman, S.P. Soil Properties Influencing Apparent Electrical Conductivity: A Review. Comput. Electron. Agric. 2005, 46, 45–70. [Google Scholar] [CrossRef]
  52. Zhou, M.; Wang, J.; Cai, L.; Fan, Y.; Zheng, Z. Laboratory Investigations on Factors Affecting Soil Electrical Resistivity and the Measurement. IEEE Trans. Ind. Appl. 2015, 51, 5358–5365. [Google Scholar] [CrossRef]
  53. Fondriest Environmental Inc. Conductivity, Salinity and Total Dissolved Solids; Fondriest Environmental Inc.: Fairborn, OH, USA, 2014. [Google Scholar]
  54. Moore, K.M. Investigation of the Hydrogeology of the Memphis Light, Gas, and Water Shaw Wellfield, Shelby County, Tennessee. Master’s Thesis, The University of Memphis, Memphis, TN, USA, 2021. [Google Scholar]
  55. Heagy, L.J.; Oldenburg, D.W. Direct Current Resistivity with Steel-Cased Wells. Geophys. J. Int. 2019, 219, 1–26. [Google Scholar] [CrossRef]
  56. Kingsbury, J.A.; Parks, W.S. Hydrogeology of the Principal Aquifers and Relation of Faults to Interaquifer Leakage in the Memphis Area, Tennessee; Water-Resources Investigations Report 93-4075; US Geological Survey: Reston, VA, USA, 1993; p. 18. [CrossRef]
  57. Gentry, R.W.; Ku, T.-L.; Luo, S.; Todd, V.; Larsen, D.; McCarthy, J. Resolving Aquifer Behavior near a Focused Recharge Feature Based upon Synoptic Wellfield Hydrogeochemical Tracer Results. J. Hydrol. 2005, 323, 387–403. [Google Scholar] [CrossRef]
  58. Hasan, K. Investigation of Modern Leakage Based on Numerical and Geochemical Modeling near a Municipal Well Field in Memphis, Tennessee. Ph.D. Dissertation, The University of Memphis, Memphis, TN, USA, 2023. [Google Scholar]
  59. Lozano-Medina, D.; Waldron, B.; Schoefernacker, S.; Antipova, A.; Villalpando-Vizcaino, R. Stories of a Water-Table: Anomalous Depressions, Aquitard Breaches and Seasonal Implications, Shelby County, Tennessee, USA. Environ. Monit. Assess. 2023, 195, 953. [Google Scholar] [CrossRef]
  60. Hossain, M.S.; Larsen, D.; Schoefernacker, S.R.; Vizcanio, R.; Hasan, M.R. Investigation of Pliocene and Pleistocene Fluvial-Terrace and Alluvial Deposits in Shelby County, Tennessee, and Their Relationship to the Shallow Aquifer System. In Geological Society of America Abstracts with Programs 159-11; The Geological Society of America: Boulder, CO, USA, 2022; Volume 54. [Google Scholar] [CrossRef]
  61. Clément, R.; Descloitres, M.; Günther, T.; Ribolzi, O.; Legchenko, A. Influence of Shallow Infiltration on Time-Lapse ERT: Experience of Advanced Interpretation. Comptes Rendus Geosci. 2009, 341, 886–898. [Google Scholar] [CrossRef]
  62. Morton, R.A.; Suter, J.R. Sequence Stratigraphy and Composition of Late Quaternary Shelf-Margin Deltas, Northern Gulf of Mexico1. Am. Assoc. Pet. Geol. Bull. 1996, 80, 505–530. [Google Scholar] [CrossRef]
  63. Castle, J.W.; Miller, R.B. Recognition and Hydrologic Significance of Passive-Margin Updip Sequences: An Example from Eocene Coastal-Plain Deposits, USA. J. Sediment. Res. 2000, 70, 1290–1301. [Google Scholar] [CrossRef]
  64. Galloway, W.E. Chapter 15 Depositional Evolution of the Gulf of Mexico Sedimentary Basin. In The Sedimentary Basins of the United States and Canada; Miall, A., Ed.; Elsevier: Amsterdam, The Netherlands, 2008; Volume 5, pp. 505–549. ISBN 1874-5997. [Google Scholar]
Water 15 04090 g002
Figure 2. Map showing (a) the location of Shelby County within the state of Tennessee; (b) map showing the locations of known or suspected breaches in Shelby County (red polygon with hatching) and Memphis aquifer recharge area (blue polygon) [15]. The black, blue, green, and red squares represent PI, AP, SF, and GC, respectively; (c) image showing the location of survey lines in SF. The black, leaf green, and brown lines represent 2.5 m, 5 m, and 10 m survey lines, respectively: (d) the location of survey lines in PI. The red and yellow lines represent lines A and B, respectively; (e) the location of the survey line at AP (blue line); and (f) the location of the survey lines in GC. The magenta and green lines represent lines 1 and 2, with 10 and 7 m electrode spacing, respectively.
Figure 2. Map showing (a) the location of Shelby County within the state of Tennessee; (b) map showing the locations of known or suspected breaches in Shelby County (red polygon with hatching) and Memphis aquifer recharge area (blue polygon) [15]. The black, blue, green, and red squares represent PI, AP, SF, and GC, respectively; (c) image showing the location of survey lines in SF. The black, leaf green, and brown lines represent 2.5 m, 5 m, and 10 m survey lines, respectively: (d) the location of survey lines in PI. The red and yellow lines represent lines A and B, respectively; (e) the location of the survey line at AP (blue line); and (f) the location of the survey lines in GC. The magenta and green lines represent lines 1 and 2, with 10 and 7 m electrode spacing, respectively.
Water 15 04090 g002
Water 15 04090 g003
Figure 3. Data-processing workflow utilized along the current study.
Figure 3. Data-processing workflow utilized along the current study.
Water 15 04090 g003
Water 15 04090 g004
Figure 4. Comparison of interpreted geology at SF with different electrode spacing: (a) IRP with 10 m electrode spacing, (b) IRP with 5 m electrode spacing, and (c) IRP with 2.5 m electrode spacing. The IRPs are complemented with superimposed lithologic logs, providing additional context for interpretation. The blue dashed line with the triangle above is the inferred water table. The black lines on each IRP indicate the inferred depth of the sand-rich, high-resistivity interval based on borehole data in wells Q-125 and Q-151 and the paleo-channel feature based on [18]. Additionally, the red dashed lines represent interpreted Eocene paleo-channel margins, while the red dotted line highlights the anomaly caused by the presence of a steel casing.
Figure 4. Comparison of interpreted geology at SF with different electrode spacing: (a) IRP with 10 m electrode spacing, (b) IRP with 5 m electrode spacing, and (c) IRP with 2.5 m electrode spacing. The IRPs are complemented with superimposed lithologic logs, providing additional context for interpretation. The blue dashed line with the triangle above is the inferred water table. The black lines on each IRP indicate the inferred depth of the sand-rich, high-resistivity interval based on borehole data in wells Q-125 and Q-151 and the paleo-channel feature based on [18]. Additionally, the red dashed lines represent interpreted Eocene paleo-channel margins, while the red dotted line highlights the anomaly caused by the presence of a steel casing.
Water 15 04090 g004
Water 15 04090 g005
Figure 5. The IRPs of (a) line 1 (10 m spacing) and (b) line 2 (7 m spacing) carried out along the GC. The solid black lines represent stratigraphic layer boundaries, and the red dashed lines show the average depth of the top of the Memphis Sand. No data for the water table is available at this site.
Figure 5. The IRPs of (a) line 1 (10 m spacing) and (b) line 2 (7 m spacing) carried out along the GC. The solid black lines represent stratigraphic layer boundaries, and the red dashed lines show the average depth of the top of the Memphis Sand. No data for the water table is available at this site.
Water 15 04090 g005
Water 15 04090 g006
Figure 6. IRP of (a) Line A and (b) Line B at PI. The gray vertical lines represent boreholes. The solid black lines represent the stratigraphic boundaries, and the solid red line indicates the channel margin. The detailed descriptions of the two boreholes are shown in Figure 7.
Figure 6. IRP of (a) Line A and (b) Line B at PI. The gray vertical lines represent boreholes. The solid black lines represent the stratigraphic boundaries, and the solid red line indicates the channel margin. The detailed descriptions of the two boreholes are shown in Figure 7.
Water 15 04090 g006
Water 15 04090 g007
Figure 7. A schematic of the lithologic profile between two boreholes drilled at PI. Fm—Formation.
Figure 7. A schematic of the lithologic profile between two boreholes drilled at PI. Fm—Formation.
Water 15 04090 g007
Water 15 04090 g008
Figure 8. The IRP of the ER survey conducted at AP. The solid gray vertical lines represent wells near the survey line, and the dashed black lines are projected layer boundaries from the geologic logs. The water table is shown with a light blue dashed line with overlying inverted triangles.
Figure 8. The IRP of the ER survey conducted at AP. The solid gray vertical lines represent wells near the survey line, and the dashed black lines are projected layer boundaries from the geologic logs. The water table is shown with a light blue dashed line with overlying inverted triangles.
Water 15 04090 g008
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Hasan, M.R.; Larsen, D.; Schoefernacker, S.; Waldron, B. Identification of Breaches in a Regional Confining Unit Using Electrical Resistivity Methods in Southwestern Tennessee, USA. Water 2023, 15, 4090. https://doi.org/10.3390/w15234090

AMA Style

Hasan MR, Larsen D, Schoefernacker S, Waldron B. Identification of Breaches in a Regional Confining Unit Using Electrical Resistivity Methods in Southwestern Tennessee, USA. Water. 2023; 15(23):4090. https://doi.org/10.3390/w15234090

Chicago/Turabian Style

Hasan, Md Rizwanul, Daniel Larsen, Scott Schoefernacker, and Brian Waldron. 2023. "Identification of Breaches in a Regional Confining Unit Using Electrical Resistivity Methods in Southwestern Tennessee, USA" Water 15, no. 23: 4090. https://doi.org/10.3390/w15234090

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop