Introduction

The hydrogeology on and near Cannon Air Force Base (AFB) in Curry County, New Mexico, was assessed to better understand preferential groundwater flow paths in the form of paleochannels. During the Triassic, streams cut channels into the geologic unit that formed the land surface at that time; remnants of these ancient streams remain in the form of paleochannels. On and near the study area, paleochannels incised the top surface of the Triassic-age Chinle Formation of Dockum Group and were subsequently filled with electrically resistive coarse-grained sediments from the overlying Tertiary-age Ogallala Formation (Rawlings, 2016), resulting in a preferential groundwater flow path in the form of a paleochannel network. A better understanding of the spatial characteristics of this preferential groundwater flow path is needed to support ongoing efforts to remediate groundwater contamination at Cannon AFB. Chandra and others (2020, p. 1) explain “paleochannels typically act as pathways for groundwater movement and provide a potential source of groundwater. Their presence can be helpful in identifying areas suitable for recharge and at times in mitigating contamination problems in afflicted regions. Thus, mapping of paleochannels is significant in the planning and management of groundwater resources.” Therefore, the U.S. Geological Survey (USGS), in cooperation with the U.S. Air Force Civil Engineer Center, used surface geophysical resistivity methods and data compiled from previous studies to better understand the spatial distribution and characteristics of the paleochannel network incised into the top of the Dockum Group.

Description of Study Area

The study area consists of about 35,000 acres on and near Cannon AFB in eastern New Mexico. Cannon AFB is approximately 7 miles southwest of Clovis, N. Mex., in the Southern High Plains physiographic province (Fenneman and Johnson, 1946; Langman and others, 2004). Cannon AFB is surrounded mostly by agricultural lands and dairy farms. The mean annual precipitation in Clovis during 1981–2020 was about 18 inches, and daily mean temperatures ranged from 78 degrees Fahrenheit in July to 39 degrees Fahrenheit in January (U.S. Climate Data, 2021). Mean annual evaporation rates in this semiarid region far exceed mean annual precipitation (Tuan and others, 1969; Robson and Banta, 1995). Cannon AFB covers approximately 3,800 acres and is host to the 27th Special Operations Group of the United States Air Force (Cannon Air Force Base, 2022) (fig. 1). Because of groundwater contamination related to historical operations at the site, an environmental restoration program is ongoing (Langman and others, 2004, 2006). Environmental restoration processes at Cannon AFB are being overseen by the Department of Defense (U.S. Government Accountability Office, 2021).

Geologic and Hydrogeologic Setting

The High Plains aquifer system is the largest aquifer in the United States and is commonly divided into three parts: the Northern High Plains aquifer, Central High Plains aquifer, and Southern High Plains aquifer (fig. 1) (Becker and others, 2002). The Southern High Plains aquifer underlies the study area. The High Plains aquifer system is commonly referred to as the “Ogallala aquifer” because the Ogallala Formation is the predominant water-bearing unit of this aquifer system (Gutentag and others, 1984).

Because the focus is on the hydrology of the area, geologic units are discussed from land surface downward (youngest to oldest), and aquifers and the geologic units that contain them are discussed together. The Quaternary-age Blackwater Draw Formation consists mostly of eolian sand deposits and overlies the Tertiary-age Ogallala Formation in the area; the Blackwater Draw Formation ranges in thickness from about 0 to 80 ft in eastern New Mexico (fig. 2; McLemore, 2001). The Blackwater Draw Formation is not considered a viable source of water in the study area. The Ogallala Formation consists of eolian sand and silt and fluvial and lacustrine sand, silt, clay, and gravel (McLemore, 2001), and ranges in thickness from about 30 to 600 ft in eastern New Mexico and western Texas (Gustavson, 1996). The Southern High Plains aquifer is the primary source of water for agriculture and public supply in the area and is primarily contained in the Ogallala Formation in eastern New Mexico (McLemore, 2001). The Ogallala Formation lies unconformably atop the upper unit of the eastward-dipping Triassic-age Chinle Formation of the Dockum Group (Dutton and Simpkins, 1986; Lucas and others, 1987). The Southern High Plains aquifer is underlain in most of the area by relatively impermeable rocks equivalent to the Chinle Formation (Cronin, 1969; Gutentag and others, 1984). The rocks that compose the Chinle Formation consist mostly of clay with some intermixed shale and silt that serve as barriers to groundwater movement (Cronin, 1969). The Chinle Formation ranges in thickness from about 0 to 400 feet (ft) in eastern New Mexico (McGowen and others, 1977). Locally, the Chinle Formation is referred to as the “red beds,” an informal name commonly used for sedimentary rocks rich in ferric oxides (Neuendorf and others, 2005). The Triassic-age Santa Rosa Formation underlies the Chinle Formation. The Chinle Formation and the Santa Rosa Formation are the only members of the Dockum Group present in the study area. In addition to the Ogallala Formation, the Santa Rosa Formation is the other water-bearing member of the High Plains aquifer system present in the study area; the sandstone and shale aquifer is contained in the Santa Rosa Formation of the Dockum Group (Gustavson and Holliday, 1985) and is as a minor source of water, mostly for agricultural uses (New Mexico Interstate Stream Commission and New Mexico Office of the State Engineer, 2016). Although in New Mexico the water-bearing unit of the Dockum Group in the study area (the Santa Rosa Formation) is known as the sandstone and shale aquifer, the water-bearing unit of the Dockum Group is known as the Dockum aquifer in Texas (Matherne and Stewart, 2012; New Mexico Interstate Stream Commission and New Mexico Office of the State Engineer, 2016).

The hydrogeologic setting on and near Cannon AFB is the complex result of the Laramide orogeny between 80 and 55 million years ago that formed the Rocky Mountains and possibly earlier orogenies as well that predate the Laramide orogeny (Gustavson and Holliday, 1985). Prior to the Laramide orogeny, eastern New Mexico and the Texas Panhandle were covered by a large sea that was part of the Western Interior Seaway, which deposited the Dockum Group sediments during the Triassic period (Gustavson and Holliday, 1985). The Dockum Group is composed of sandstone with interbedded shales grading upward to a shaly sandstone or siltstone and clay (Luckey and Becker, 1999). When the Laramide orogeny began and the Western Interior Seaway regressed, the area on and near the present-day Cannon AFB was traversed by large rivers that had their headwaters in the Rocky Mountains and predominately flowed in an east-southeast direction. These rivers downcut into the Chinle Formation that forms the upper part of the Dockum Group until the late Miocene epoch (Gustavson and Holliday, 1985) when the incised channels began to fill with fluvial sediments such as sand and gravel originating from the Rocky Mountain headwaters; subsequent to channel incision and streambed sediment deposition, sand and gravel deposits from complex ancestral alluvial fan systems were deposited and then reworked to form the overlying Ogallala Formation (Knowles and others, 1984).

Data Collection, Compilation, and Processing Methods

Resistivity measurements are commonly used to interpolate geologic and hydrogeologic properties between borehole locations or in areas where no borehole data exist. Data from surface geophysical resistivity methods were used to fill data gaps pertaining to the spatial distribution of paleochannels at Cannon AFB. Surface geophysical resistivity measurements in the form of time-domain electromagnetic (TDEM) soundings made by the USGS were published as a companion data release to this report (Payne and Teeple, 2020). These TDEM soundings were used in conjunction with data published in AECOM (2020) and borehole data compiled from NMWRRS (2021) to prepare an updated map of the top of the Dockum Group that includes the location and characteristics of paleochannels incised into the top of the Dockum Group (Chinle Formation).

Hydrogeologic Unit Interpretation

In order to determine the top of Dockum Group (Chinle Formation) (fig. 3), the tops and bases of geologic units (and their hydrogeologic-unit equivalents), hereinafter referred to as “borehole picks,” were first determined by analyzing previously published and newly collected lithologic descriptions and borehole geophysical logs. A total of 149 borehole picks were obtained from AECOM (2020), along with 72 borehole picks from NMWRRS (2021) and 43 borehole picks from newly collected TDEM soundings. The data were gridded using Oasis Montaj v. 9.8.1 (Seequent, 2020a) where a kriging method using an exponential variogram model trended to the southeast. Kriging is a geostatistical method that determines the most probable value at each grid node in a 50- by 50-m grid (about 164 ft between nodes in the x and y directions) for this study based on a statistical analysis of the entire dataset (Isaaks and Srivastava, 1989). Variance maps developed during the kriging process were used to evaluate the uncertainty in hydrogeologic unit surface grids in the planning of additional data-collection tasks. Generally, as the distance between data points became greater, the correlation between the data points decreased, and uncertainty in areas between the data points increased (Isaaks and Srivastava, 1989). Additional information on kriging is available in Isaaks and Srivastava (1989). The contour lines shown in figure 3 were created by using the quick contour method in Oasis Montaj v. 9.8.1 (Seequent, 2020a), are based on available data, and were not adjusted manually. Figure 3 depicts the updated general northwest-to-southeast orientation of the top surface of the Dockum Group (Chinle Formation) that underlies Cannon AFB. The blue-shaded areas in figure 3 correspond to the general location of the paleochannel network, and the result obtained is consistent with the results from previous studies (AECOM, 2020).

Previous base-of-aquifer maps generated for the area have shown areas with water levels below the proposed base of the aquifer; therefore, the area should have been dry. This modified grid compares much better to water level data in the area and has no areas that do not correlate with the other datasets. The minimum saturated thickness in a well that has water level data is approximately 10 ft.

A map of the grid error (variance) derived from the kriging method was developed to depict where possible data gaps may still exist (fig. 4). The updated map of the top of Dockum Group has many areas of uncertainty greater than 20 ft (fig. 4) because there are not enough data for the gridding process to reliably determine a value. However, this interpretation of the altitude of the top of the Dockum Group represents a substantial improvement in data resolution compared to previous studies. The resolution over the entire study area improved because of the additional borehole data, and the resolution along the large paleochannel trending northwest-to-southeast across Cannon AFB is greatly improved by the addition of new, tightly spaced TDEM soundings. Many data gaps are still indicated (fig. 4), and additional data collection is needed to further improve the accuracy with which the top of the Dockum Group is depicted in areas where data remain sparse. Areas within Cannon AFB that would benefit most from additional data collection are in locations where infrastructure hindered the collection of additional TDEM soundings. The depiction of the top of the Dockum Group in areas east and south of the Cannon AFB may benefit from additional TDEM soundings, including areas where difficulty accessing private land hindered data collection.

Paleochannel Analysis

Two methodologies were used to evaluate paleochannels in the top of the Dockum Group across the study area: (1) trend-removal grid analysis and (2) analysis with Esri’s ArcMap Hydrology toolset. Trend-removal grids were developed by removing the third-order regional trend from all the data points within the top of Dockum Group (Chinle Formation) surface grids (Seequent, 2020b). Essentially, this technique removes the regional dip of the layer prior to the process of creating surface grids. The resulting trend-removal grid (fig. 5) shows changes in the top of the Dockum Group after the regional trend was removed. Relatively low altitude values represent low-altitude data points in the top of Dockum Group (Chinle Formation) surface grid, and continuous stretches of these low-altitude values can be indicative of paleochannels. Figure 5 shows the same general northwest-to-southeast trend of paleochannel network across Cannon AFB, but with the regional trend removed, it highlights relatively deeper portions of the paleochannels. The areas in figure 5 with darker blues indicate the portions of paleochannels that are substantially deeper (more than 20 ft) relative to the main paleochannel. These deeper areas of the paleochannel indicate portions with greater saturated thickness.

The second paleochannel analysis, performed using Esri’s ArcMap Hydrology toolset, treated the surface grid representing the top of the Dockum Group as a digital surface elevation model and simulated a topographic watershed across the surface to identify the location and spatial characteristics of the paleochannel network. The toolset was used to determine the flow direction, flow accumulation, and stream order (Esri, 2021). Ultimately, the toolset simulates the movement of water across a land surface, which for this study corresponds to the top of the Dockum Group surface grid (fig. 3).

Concentric “bullseye” shapes were present in the top of Dockum Group surface contours that were not smoothed manually. For the toolset to function properly, these “bullseyes” (which act as sinks) had to be filled to render the raster hydrologically correct. Next, a flow-direction grid was created to find directional flow between cells and then used as input to the flow-accumulation tool, where a threshold lower than the default value was chosen to identify paleochannel flow paths (Esri, 2021). This flow-accumulation grid was then converted to vector format, where artifacts of the process were removed manually. The final product is a map depicting flow paths (fig. 5) based on the interpolated and smoothed top of Dockum Group altitudes. Data from previous USGS studies were used to check the results from the two methodologies used to evaluate paleochannels in the top of the Dockum Group across the study area.

These two paleochannel analysis techniques show groundwater flow direction as well as areas with the deepest saturated thickness. Hydrologically, these techniques show where aquifer storage is highest (in the deepest parts of the paleochannel network), as well as the spatial distribution of preferential groundwater flow paths (the paleochannels [fig. 5]). The analyses indicate a large paleochannel trending to the southeast with smaller channels feeding in from the west. Areas where groundwater management could be more beneficial are indicated by locations where these flow lines intersect the deeper parts of the paleochannel.

Summary

The hydrogeology on and near Cannon Air Force Base (AFB) was assessed to gain a better understanding of preferential groundwater flow paths through paleochannels. The Southern High Plains aquifer is contained in the Tertiary-age Ogallala Formation in the part of eastern New Mexico where Cannon AFB is located. The Triassic-age Chinle Formation of the Dockum Group underlies the unconfined Southern High Plains aquifer. The Southern High Plains aquifer is the primary source of water for agriculture and public supply in the area, with the sandstone and shale aquifer serving as a minor source of water mostly for agricultural uses. In and near the study area, paleochannels incised the top surface of the Dockum Group (Chinle Formation) and subsequently filled in with electrically resistive coarse-grained sediments of the overlying Ogallala Formation, resulting in a preferential groundwater flow path in the form of a paleochannel network. A better understanding of the spatial characteristics of this preferential groundwater flow path is needed to support ongoing efforts to remediate groundwater contamination at Cannon AFB. Therefore, the U.S. Geological Survey, in cooperation with the U.S. Air Force Civil Engineer Center, used surface geophysical resistivity methods and data compiled from previous studies to better understand the spatial distribution and characteristics of the paleochannel network incised into the top of the Dockum Group.

Previous studies have shown with increasing resolution how paleochannels incised the top of the Dockum Group, but limited borehole data on and near Cannon AFB continued to make accurately mapping the top of Dockum Group challenging. For this study, surface geophysical resistivity measurements in the form of time-domain electromagnetic (TDEM) soundings made by the U.S. Geological Survey were used in conjunction with data previously published by Architecture, Engineering, Construction, Operations, and Management and borehole data compiled from the New Mexico Water Rights Reporting System database to prepare an updated map of the top of the Dockum Group that includes the location and characteristics of paleochannels incised into the top of the Dockum Group (Chinle Formation). A total of 149 borehole picks were obtained from previous studies, along with 72 additional borehole picks from the New Mexico Water Rights Reporting System database and 43 borehole picks from newly collected TDEM soundings. The data were gridded and contoured using Oasis Montaj v. 9.8.1 where a kriging method using an exponential variogram model trended to the southeast was used. A map of the grid error (variance) derived from the kriging method was developed to depict where possible data gaps may still exist. The updated map of the top of Dockum Group has many areas of uncertainty greater than 20 feet because there are not enough data for the gridding process to reliably determine a value. However, this interpretation of the altitude of the top of the Dockum Group represents a substantial improvement in data resolution compared to previous studies. The resolution over the entire study area improved with the additional borehole data, and the resolution along the large paleochannel trending northwest-to-southeast across Cannon AFB is greatly improved by the addition of new, tightly spaced TDEM soundings. Many data gaps remain, and additional data collection is needed to further improve the accuracy with which the top of the Dockum Group is depicted in areas where data remain sparse.

Two methodologies were used to evaluate paleochannels in the top of the Dockum Group across the study area: (1) trend-removal grid analysis and (2) analysis with Esri’s ArcMap Hydrology toolset. Trend-removal grids were developed by removing the third-order regional trend from all the point values within the top of Dockum Group (Chinle Formation) surface grids. The resulting trend-removal grid shows changes in the top of the Dockum Group after the regional trend was removed. Relatively low values represent low-altitude points in the top of Dockum Group (Chinle Formation) surface grid and continuous stretches of these low-altitude points can be indicative of paleochannels. The grid shows the same general northwest-to-southeast trend of paleochannel network across Cannon AFB, but with the regional trend removed, it highlights relatively deeper portions of the paleochannels (more than 20 feet relative to the main paleochannel). These deeper areas of the paleochannel indicate areas with more saturated thickness. Esri’s ArcMap Hydrology toolset was used to determine the flow direction, flow accumulation, and stream order of paleochannels. A flow-direction grid was first created to find directional flow between cells, then used as input to the flow-accumulation tool, where a threshold lower than the default value was chosen to identify paleochannel flow paths. The final product is a map depicting flow paths based on the interpolated and smoothed top of Dockum Group altitudes. These two paleochannel analysis techniques show groundwater flow direction as well as areas with the deepest saturated thickness. Hydrologically, these techniques show where aquifer storage is highest (in the deepest parts of the paleochannel network), as well as the spatial distribution of preferential groundwater flow paths (the paleochannels). The analyses indicate a large paleochannel trending to the southeast with smaller channels feeding in from the west. Areas where groundwater management could be more beneficial are indicated by locations where these flow lines intersect the deeper parts of the paleochannel.