Geologic History

The geologic history of the Arbuckle Mountains encompasses more than a billion years, from Proterozoic igneous and metamorphic rocks to Quaternary alluvial deposits. There are four main phases of geologic history, including tectonics and sedimentation, that formed and shaped the Arbuckle Mountains of today: (1) an initial splitting apart of tectonic plates (rifting) during the Early and Middle Cambrian Epochs, (2) deposition and subsidence during the Late Cambrian Epoch through the Mississippian Subperiod, (3) uplift and deformation during the Pennsylvanian Subperiod, and (4) erosion and post-Pennsylvanian Subperiod tilting (Johnson, 1991). Before the first phase of rifting, during the Proterozoic Eon, the study area was underlain primarily by granites and gneisses. About 1.3 billion years ago, during the Proterozoic Eon, igneous dikes were intruded into the surrounding rock, which was the first evidence of crustal weakness that would later alter the geology of the area.

In the Early and Middle Cambrian Epochs, the first phase of crustal deformation occurred, with rifting that caused the development of major normal faults along the margins and more igneous activity. As these igneous rocks cooled in the rift zone, the land surface began to subside, creating a trough that would subsequently be filled with sedimentary rocks.

The Late Cambrian Epoch through the Mississippian Subperiod is marked by the second main phase of geologic history in the region. During this time, shallow seas covered much of the midcontinent of what is now North America, which caused deposition of carbonate sediments that would later become the Arbuckle Group that was deposited during the Late Cambrian Epoch through the Early Ordovician Epoch. Some of these sediments were deposited in the sedimentary trough, which is commonly referred to today as the Southern Oklahoma Aulacogen (Johnson and others, 1989; Chase and others, 2022). Sedimentary rocks that formed in the Southern Oklahoma Aulacogen are much thicker than the surrounding rock, with more than 17,000 feet (ft) of rock accumulation, relative to the surrounding continental shelf, which accumulated about 6,500 ft of rock during the same time period (Ham, 1973). At the end of the Early Ordovician Epoch the sea level decreased enough to expose these carbonate sediments to meteoric waters, resulting in the dolotomization of some rocks in the Arbuckle Group (Lynch and Al-Shaieb, 1991; Denison, 1997). Dolomitization is a process where the carbonate mineral dolomite is formed when magnesium ions replace calcium ions in calcite. This second phase of geologic history also saw the deposition of the Simpson Group of Middle to Late Ordovician age on top of the rocks of the Arbuckle Group. The Simpson Group is composed of marine-shelf carbonates, shales, and quartz sandstones, indicating a shallow-sea depositional environment (Johnson, 1991).

The third phase of geologic history in the Arbuckle Mountains began in the Early Pennsylvanian Epoch and was marked by uplift and deformation, including the formation of the Hunton Anticline in the study area. The Late Pennsylvanian Epoch brought intense mountain building along the margins of the Southern Oklahoma Aulacogen, which resulted in tight folding and high angle thrust faulting of these rocks (Christenson and others, 2011). This mountain building event is thought to have been caused by a major plate collision between the North American plate and Gondwana or another smaller plate during the Late Pennsylvanian Epoch (Perry, 1989).

Finally, the fourth phase of geologic history in the study area is characterized by additional deposition and erosion. After the Late Pennsylvanian Epoch mountain-building event, sediment deposits of Late Pennsylvanian age covered the Arbuckle Mountains. Red beds and evaporites of Permian age were deposited in basins, followed by shallow-sea sand and carbonate deposits of Cretaceous age (Johnson and others, 1989). These shallow seas also caused erosion of the Arbuckle Mountains, which were further flattened by fluvial erosion during the Cretaceous Period (Donovan, 1991). Further erosion and tilting in the study area was due to the uplift of the Rocky Mountains hundreds of miles west of the Arbuckle Mountains during the Laramide orogeny (English and Johnston, 2004). Alluvial and terrace sedimentation of Quaternary age deposited along streams and rivers round out this period of deposition and erosion in the study area (Johnson and others, 1989). A more complete geologic history of the Arbuckle Mountains was detailed by Christenson and others (2011).

Structural Geology

The geologic units that contain the eastern part of the Arbuckle-Simpson aquifer, which is the focus of this study, are shaped primarily by the Hunton Anticline, but include other structural features, including the Belton and Clarita Anticlines, the Sulphur Syncline, and the Lawrence Uplift (fig. 2). The Hunton Anticline is a broad anticlinal fold that exposed the Early Ordovician-age West Spring Creek and Kindblade Formations of the Arbuckle Group to the surface in the central part of the eastern part of the Arbuckle-Simpson aquifer (Johnson, 1990). The northwestern arm of the Hunton Anticline dips gently (less than 20 degrees; Christenson and others, 2011) westward, while the southeastern arm dips gently to the east. The Hunton Anticline is bounded by the Lawrence Uplift to the north, the Franks Fault Zone and the Clarita Fault to the northeast/east, and the Sulphur Fault Zone to the south (Christenson and others, 2011). The Sulphur Fault Zone consists of major northwest-southeast-trending faults, which were first displaced during the formation of basement rocks and were reactivated multiple times during Paleozoic Era rifting and orogeny (Harlton, 1966; Denison, 1995). The Sulphur Fault Zone roughly coincides with the northern edge of the Sulphur Syncline, which is bounded to the south by the South Sulphur Fault. Rocks from the Simpson Group were folded to form the Sulphur Syncline, which terminates to the southeast of the study area in the east and at approximately the Chickasaw National Recreation Area to the west (fig. 2). Results from a geophysical gravity study suggest that the Sulphur Syncline might be a graben rather than a syncline, but the name remains the same (Cates, 1989; Scheirer and Hosford Scheirer, 2006).

South of the Sulphur Syncline lies the Belton Anticline, which makes up the southern portion of the eastern part of the Arbuckle-Simpson aquifer study area. The Belton Anticline is a northwest-plunging folded fault block that is structurally higher than the Hunton Anticline (Christenson and others, 2011). These structurally high areas have been eroded, which has exposed the Early Ordovician-age Cool Creek and McKenzie Hill Formations of the Arbuckle Group to the surface. The Mill Creek Fault marks the southern boundary of the eastern part of the Arbuckle-Simpson aquifer study area. This fault runs nearly parallel to the Sulphur Fault Zone, with a northwest-southeast trend, and is nearly vertical, based on gravity data collected near the Chickasaw National Recreation Area (Christenson and others, 2011). Several other small faults in a range of orientations mark the subsurface of the study area, further enhancing disjointed and preferential groundwater-flow paths (Fairchild and others, 1990; Riley, 2004; Scheirer and Hosford Scheirer, 2006; Sample, 2008; Kennedy and others, 2009; Halihan and others, 2009a; Smith and others, 2009; Young and others, 2009; Ramachandran and others, 2012). Detailed visualizations of the structural geology of the eastern part of the Arbuckle-Simpson aquifer area, including geologic unit outcroppings, major faults, structural features, and cross sections, can be found in figures 4, 6, and 7 of Christenson and others (2011).

Hydrogeologic Units

The geologic units of the Arbuckle Mountains, which include the units that contain the Arbuckle-Simpson aquifer, are basement rocks (Cambrian rhyolites and Proterozoic granites and gneisses), the Timbered Hills Group of Late Cambrian age, the Arbuckle Group, and the Simpson Group.

The basement rocks of the Arbuckle Mountains were formed in the Ectasian Period of the Mesoproterozoic Era and are the oldest rocks in the study area at 1.35 to 1.4 billion years old (Ham and others, 1964). They consist of igneous and metamorphic rocks and include the Tishomingo Granite, Troy Granite, and unnamed granodiorite and granitic gneiss. In the eastern part of the Arbuckle-Simpson aquifer study area, these basement rocks are only exposed at the surface as the core of the Belton Anticline in the southern part of the study area (Ham and others, 1964; Denison, 1973). These igneous and metamorphic rocks are cut by several Proterozoic- and Cambrian-age dikes and extrusive pyroclastic rocks. No high-yield water wells are completed in these igneous and metamorphic rocks because they are believed to have very low hydraulic conductivity due to their crystalline structure (Christenson and others, 2011). As a result, these basement rocks form a lower confining unit to the eastern part of the Arbuckle-Simpson aquifer with depths ranging from 3,100 to 4,600 ft below land surface (bls) (Campbell and Weber, 2006).

The Arbuckle Group overlies the Timbered Hills Group and consists of Late Cambrian- to Early Ordovician-age carbonate rocks. The Arbuckle Group is approximately 3,000 ft thick in most of the eastern part of the Arbuckle-Simpson aquifer study area but has been eroded away over parts of the Belton Anticline (Christenson and others, 2011). The Arbuckle Group is divided into eight formations, which include the Fort Sill Limestone and Royer Dolomite of Late Cambrian age, the Signal Mountain Formation of Early Ordovocian age, and the Butterly Dolomite, McKenzie Hill Formation, Cool Creek Formation, Kindblade Formation, and West Spring Creek Formation of Early Ordovocian age, from oldest to youngest (fig. 3). Dolostones are the most common type of carbonate rock in the Arbuckle Group in the eastern part of the Arbuckle-Simpson aquifer (Ham, 1945). Numerous unconformities throughout the Arbuckle Group are evidence that the carbonate rocks were exposed to weathering, encouraging the development of karst features. Surface exposures of the Arbuckle Group, as well as cores of the subsurface, show examples of paleokarst features including dissolution in fractures and cavities, vuggy porosity, and collapse breccias. These paleokarst features increase the porosity of rocks in the Arbuckle Group, also increasing their permeability (Lynch and Al-Shaieb, 1991). The Arbuckle Group is the larger (by both area and thickness) of the two lithostratigraphic groups that contain the Arbuckle-Simpson aquifer. The portion of the Arbuckle-Simpson aquifer contained in the Arbuckle Group is also more productive than the portion contained in the Simpson Group (Christenson and others, 2011) because of its intercrystalline porosity and because of the numerous fractures, solution channels, and cavities it contains (Fairchild and others, 1990). Wells completed in the portion of the Arbuckle-Simpson aquifer contained in the Arbuckle Group typically yield 200 to 500 gallons per minute (gal/min), with some deeper wells reported to yield up to 2,500 gal/min (Fairchild and others, 1990; Christenson and others, 2011).

The Simpson Group is the younger of the lithostratigraphic groups contained in the Arbuckle-Simpson aquifer. The Simpson Group is generally less than 1,000 ft thick in the eastern part of the Arbuckle-Simpson aquifer and is exposed at the surface in about one-third of the total aquifer area (145 mi²; Christenson and others, 2011), typically along the edges of major anticlines and other structurally low areas. Erosion over the structurally higher locations in the aquifer has removed the Simpson Group from those areas (Ham, 1973). The largest outcrop of the Simpson Group in the eastern part of the Arbuckle-Simpson aquifer is on the eastern flank of the Hunton Anticline, with another exposure on the Sulphur Syncline.

The Simpson Group was deposited during a time of sea level fluctuations, which resulted in the formation of porous quartzose sandstones interbedded with limestones, dolostones, and greenish-gray shales. The Simpson Group consists of five formations: the Joins, Oil Creek, and McLish Formations of Middle Ordovician age, and the Tulip Creek and Bromide Formations of Late Ordovician age (fig. 3). Because the Joins and Tulip Creek Formations are either very thin or absent in the eastern part of the Arbuckle-Simpson aquifer (Christenson and others, 2011), they will not be discussed further. The most well-developed formations in the eastern part of the Arbuckle-Simpson aquifer are the basal Oil Creek and McLish Formations, which have sandstones with thicknesses of up to 400 and 165 ft, respectively (Ham, 1945; Denison, 1997). These thick beds of uncemented quartz sandstones are mined locally to produce glass, and they store the majority of the water in the part of the eastern part of the Arbuckle-Simpson aquifer contained in the Simpson Group. Wells completed in the eastern part of the Arbuckle-Simpson aquifer contained in the Simpson Group typically yield 100 to 200 gal/min (Fairchild and others, 1990).

Where the top of the eastern part of the Arbuckle-Simpson aquifer is not exposed at the surface, it is confined above by younger rocks of various ages deposited after the Simpson Group (fig. 3). This unnamed upper confining unit confines the eastern part of the Arbuckle-Simpson aquifer on the western edge of the Hunton Anticline, near Sulphur, Okla., and Chickasaw National Recreation Area. This unnamed upper confining unit consists primarily of conglomerate with some sandstone, shale, and minor nodular limestone and lies unconformably over the Arbuckle and Simpson Groups.

Quaternary alluvial deposits are the youngest sediment deposits in the study area. These deposits, consisting of unconsolidated gravel, sand, silt, and clay, are primarily found along larger streams in the study area and are typically very thin and poorly defined. The nature of the Quaternary alluvial deposits is such that they likely would not have the properties of a confining unit; however, these deposits are also not expected to have appreciable hydrologic interaction with the eastern part of the Arbuckle-Simpson aquifer and were, therefore, included in the unnamed upper confining unit for simplicity.

Base Flow

Base-flow separation is the method of separating the base-flow or groundwater component from the runoff component to determine the percentage of each that compose the total streamflow. Base-flow separation methods are based on the assumptions that water discharging an aquifer to a stream is continuous. Over time, changes in the ratio of base flow to total streamflow may indicate changes in groundwater storage or availability. Base-flow separation was completed by using the PART program of the USGS Groundwater Toolbox (Barlow and others, 2015). The PART method uses streamflow partitioning to estimate daily base flow from the streamflow record and is based on the antecedent streamflow recession (Rutledge, 1998; fig. 6). The PART method is used for the analysis of the groundwater-flow system of a basin for which a streamflow-gaging station at the downstream end can be considered the only point of outflow. One assumption of the PART method is that the area of the contributing groundwater-flow system is equal to the drainage area of the streamflow-gaging station for the purpose of expressing flow in units of specific discharge (length per time). To use the PART method, regulation and diversion of flow should be negligible and the drainage-basin area should be less than 500 mi² (Rutledge, 1998). For this report, negligible regulation for a streamflow-gaging station is defined as having less than 20 percent of the drainage area upstream from a streamflow-gaging station controlled by dams, floodwater-retarding structures, or other human modifications of streamflow. Analysis of hydrographs using PART results in periods of time (annual, monthly, seasonal) with base-flow and runoff portions of streamflow for which base-flow percentages can be calculated. Base-flow separation during Phase 1 and Phase 2 was done by using the PART hydrograph-separation method included in the USGS Groundwater Toolbox (Barlow and others, 2015). For the PART hydrograph-separation computations, the subsurface watershed areas used were from table 8, page 39, in Christenson and others (2011).

Base-flow separation was completed for the Phase 1 studies for calendar years 2004–08 for the streamflow records collected at the Blue River streamgage and Pennington Creek near Reagan gage, and for calendar years 2007–08 for the streamflow records collected at USGS streamgage 07331200 Mill Creek near Mill Creek, Okla. (hereinafter referred to as the "Mill Creek near Mill Creek gage"). Mean annual base flow was 74.2 percent of the total streamflow measured at the Blue River streamgage (fig. 6). Mean annual base flow computed at the Pennington Creek near Reagan gage was 82.8 percent of the total streamflow measured in Pennington Creek. Mean annual base flow computed at the Mill Creek near Mill Creek gage was 52.5 percent of the total streamflow measured in Mill Creek. The streamflow measured at the Byrds Mill Spring gage was 100 percent base flow because it consisted entirely of groundwater issuing from Byrds Mill Spring with no surface-water component. During 1990–2005, streamflow at the Byrds Mill Spring gage averaged 18.5 cubic feet per second (ft³/s).

Base-flow separation was completed for Phase 2 for calendar years 2018–22 at the Blue River streamgage, the Pennington Creek near Reagan gage, and the Mill Creek near Mill Creek gage. Mean annual base flow at the Blue River streamgage was 70.3 percent of the total streamflow (fig. 6). Mean annual base flow at the Pennington Creek near Reagan gage was 79.7 percent of the total streamflow. Mean annual base flow at the Mill Creek near Mill Creek gage was 61.1 percent of the total streamflow. Base flow increased from Phase 1 to Phase 2 at all three streamgages. However, the base-flow portion of streamflow decreased from Phase 1 to Phase 2 at two of the three streamgages analyzed (fig. 6).

One issue with attempting to compute the base-flow portion of streamflow is the influence of anthropogenic discharges upstream, such as stream augmentation from producing mines (described in the “Consumptive Water Use at Producing Mines” section of this report). These discharges affect streamflows measured and analyses of streamflows for interpretation of groundwater and surface-water interactions and base-flow analyses (fig. 7). The increases in the base-flow portion of streamflow and total streamflow during Phase 2 (fig. 6C) could be related to these upstream discharges.

Net Streamflow Gains and Losses

Net gains and losses in streamflow along stream segments can be quantified by using discrete discharge measurements made during base-flow conditions, commonly referred to as synoptic base-flow (seepage-run) measurements. Locations of these measurements can be mapped to illustrate interactions with the underlying aquifer during base-flow conditions. Discharge was measured at 40 stream locations at USGS stations, including several USGS streamgages, across the land surface overlying the aquifer in each stream main stem and selected tributaries (table 6; figs. 8–9). The relative net gain or loss of groundwater from or to the aquifer for each stream segment between two measurements was calculated (figs. 8–9). To ensure base-flow conditions were being captured, with as little runoff as possible, streamflow discharge was measured during January 24–31, 2022 (fig. 8), and February 28–March 1, 2023 (fig. 9), when evapotranspiration and groundwater withdrawals were considered minimal and after runoff from any precipitation events had dissipated (streamflow hydrographs in and near the study area were analyzed to ensure runoff had dissipated). Fewer sites were measured in 2023 due to lack of access to sites.

Discrete discharge measurements, referred to as synoptic base-flow measurements (seepage runs), were made at several sites along streams during a period of relatively low precipitation in a short period of time to determine base flow. Synoptic base-flow measurements were collected by using the methods of Rantz and others (1982) and Turnipseed and Sauer (2010). Each streamflow measurement was assigned a quality rating by the field technician. The following ratings were assigned to measurements based on USGS guidelines (Turnipseed and Sauer, 2010): excellent, measured discharge was within 2 percent of actual discharge; good, measured discharge was within 5 percent of the actual discharge; fair, measured discharge was within 10 percent of the actual discharge; poor and unspecified, measured discharge was assumed to be within 8 percent of actual discharge (table 6). Gaining and losing segments were determined by calculating the difference in seepage (streamflow discharge) measurements at each end of a segment. To calculate a rate of gain or loss per mile along the segment, the difference in discharge was divided by the stream length between the upstream and downstream measurement. Base flow increases in the downstream direction in a gaining stream as water seeps into the stream from the aquifer, whereas base flow decreases in the downstream direction in a losing stream as water seeps out of the stream to the aquifer (Winter and others, 1998). Tributary inflow was accounted for in these measurements by subtracting tributary discharge from the upstream measurement.

The 2022 seepage measurements indicated net gaining stream reaches along portions of Rock Creek, Mill Creek, Pennington Creek, and the Blue River (fig. 8). Net gaining stream reaches indicate that groundwater from the Arbuckle-Simpson aquifer is flowing into streambeds and maintaining streamflows. Upstream reaches of the Blue River to the north of site AM-18 were net losing; streamflows ranged from 0.00 to 4.28 ft³/s in these reaches, indicating net seepage losses from the stream into the Arbuckle-Simpson aquifer. Streamflows measured between sites AM-18 and AM-19 indicated a gain of 2.33 ft³/s per mile. Streamflow measured at the Blue River streamgage (site AG-12) and the measurement upstream at site AM-25 indicated a gain of 7.63 ft³/s per mile. There was no flow in the upper reach of Pennington Creek upstream from site AG-06 or between the next two downstream seepage measurement sites AM-11 and AM-12. There was no flow in Mill Creek upstream from sites AM-07 and AG-03, but there was streamflow between sites AG-03 and AG-04, and in this reach streamflow increased by 0.27–0.28 ft³/s per mile. Delaware Creek lost streamflow between sites AM-26 and AG-13. Little Blue Creek, a tributary to the Blue River, gained streamflow of 0.26 ft³/s per mile between sites AM-21 and AM-20.

The 2023 seepage measurements indicated that more stream segments were gaining than during the 2022 seepage measurements on Mill Creek, Pennington Creek, Blue River, Little Blue Creek, and Delaware Creek (fig. 9). In addition, streamflows computed at USGS streamgages in the study area were generally greater during February 28–March 1, 2023, than during January 24–31, 2022. For example, streamflow values at sites AG-09 and AG-10 during 2022 were both 0.0 ft³/s, but streamflow values during 2023 at those streamgages were 12.7 and 11.5 ft³/s, respectively.

Continuous Springflow Monitoring

Discharge was continuously monitored at USGS streamgages at four major springs in the study area: Antelope Spring, unnamed Blue River spring, Byrds Mill Spring, and Sheep Creek Spring (fig. 2). Byrds Mill Spring near Fittstown, Okla. (07334200; table 1–1 in appendix 1) was monitored during 1959–2017 at the Byrds Mill Spring gage; data were used in the Phase 1 analyses, but this gage was discontinued prior to the initiation of the Phase 2.

Seasonal patterns were observed in spring discharge. Spring discharge was generally lower in summer and winter when there is typically less precipitation compared to fall and spring, and evaporation losses reach their annual peak in the summer, which also causes spring discharge to decrease relatively more than during other seasons. For all seasons, spring discharges were generally highest during the spring because seasonal precipitation amounts are highest and because evapotranspiration rates in the spring are relatively low compared to those in the summer.

Continuous spring discharge hydrographs from the Antelope Spring gage, Byrds Mill Spring gage (USGS streamgage 07334200), and Sheep Creek Spring gage were analyzed for the highest discharge periods during 2015–16 to identify flow regimes of this karst aquifer. Linear regression equations (regression curves) for spring discharge and time were examined for breaks (inflection points) in the recession slope, which are indicative of a change from one flow regime to another within the karst continuum (Otero, 2007). Diffuse flow is the amount of water that flows through the rock matrix (Kresic, 2013). Conduit flow is the amount of water that flows through rock fractures or conduits, which are interconnected solution cavities where the length is disproportionally larger than the height or width (Kresic, 2013). Recession curves were classified as different flow regimes based on their decay coefficient (α) in Milanović’s (1976) equation:

If the decay coefficient is 0.18, then conduit flow is the predominant flow type. If the decay coefficient is 0.09, then flow type consists of fracture flow or a mixture of conduit flow and diffuse flow (with conduit dominant). If the decay coefficient is 0.02, then flow type consists of fracture flow or a mixture of conduit flow and diffuse flow (with diffuse dominant). If the decay coefficient is 0.008, the diffuse flow is the predominant flow type (Milanović, 1976). As indicated by the recession curves at the highest discharge periods, Antelope Spring yielded a higher decay coefficient than during the other periods of recession and exhibited more conduit flow than diffuse flow, followed by greater portions of mixed and diffuse flow (fig. 10A). This higher decay coefficient indicates that during periods of high-discharge recession, the groundwater system supplying Antelope Spring might include upper layers of the aquifer dominated by fractures and conduits, with a middle and lower section of the aquifer dominated by fractures, primary porosity, or both.

As indicated by recession curves for the highest discharge periods for the Byrds Mill Spring and Sheep Creek Spring gages, decay coefficients are relatively lower and flow is mostly diffuse compared to Antelope Spring (fig. 10B, C). These lower decay coefficients indicate that the groundwater system near Byrds Mill and Sheep Creek Springs, near the upper levels of the formation, is likely dominated by diffuse drainage through primary porosity.

Discrete Springflow Measurements

Discrete springflow measurements were made during 2022–23 at springs in the study area (table 1–2 in appendix 1) to document current springflows (at the time when the measurements were made) across the eastern part of the Arbuckle-Simpson aquifer for the Phase 2 study and to better understand changes in springflow over time (table 7). As explained in the “Introduction” section of this report, changes in the springflow over time across the aquifer could be an indicator of change in aquifer water storage amounts. Spring discharges were not measured and documented specifically for the Phase 1 studies, but historical spring discharges from the USGS NWIS database (1954–2017) (USGS, 2024a) were analyzed by the OWRB to identify which springs to include for the well-spacing rules for sensitive sole-source groundwater basins. Spring discharges for Phase 2 were measured during 2018–23, with some multiple measurements for the Phase 2 period of record specified in table 8 to indicate the period for which the Phase 2 mean discharge was calculated. The OWRB well-spacing rules are dependent on springs flowing greater than 50 gal/min (0.11 ft³/s) and 500 gal/min (1.11 ft³/s). Springflows as determined from historical discharge measurements were compared to Phase 2 discharge measurements (table 7). Eleven of the 17 spring sites had a decrease in discharge from the historical period to Phase 2 (table 7). Additional spring locations were documented, and discharge was measured for Phase 2 (table 1–2, appendix 1).

Potentiometric Surfaces

A potentiometric surface is a surface of equal hydraulic head or potential, typically depicted by a map of equipotential lines such as water-table elevations (Sharp, 2024) and thus represents a snapshot of groundwater levels (commonly referred to as the water table) across the aquifer for a specific point in time. Specific points on potentiometric-surface maps are roughly equivalent to the level to which water will naturally rise in a tightly cased well. Potentiometric-surface maps can be used to help identify groundwater-flow directions, delineate subsurface groundwater basins, and estimate changes in groundwater storage by comparing potentiometric surfaces from different time periods (Driscoll, 1986). The altitudes of potentiometric surfaces of the Arbuckle-Simpson aquifer were calculated by collecting groundwater-level measurements at 57 groundwater wells across the study area in 2022 (fig. 12) and 56 wells in 2023 (fig. 13). These measurements were collected during base-flow conditions in mid-February 2022 and 2023, when evapotranspiration rates and groundwater use are less compared to other months.

A potentiometric surface is a theoretical topographical surface which represents the fluid potential of groundwater within an aquifer. Potentiometric surfaces were constructed in ArcGIS Pro by combining synoptic water-level data with a digital elevation model (DEM) and generating contours around the derived static head measurements (depth to water, in feet below land surface datum). Contours were adjusted to remedy irregularities resulting from automated contour generation and known inconsistencies in the study area. The resultant map displays a surface that can be used to interpret the direction groundwater will flow throughout a given aquifer.

Groundwater-level data were used to create two potentiometric-surface maps over the main area of the eastern part of the Arbuckle-Simpson aquifer (figs. 12–13). Both maps show a decrease in water-level altitude from the northwest to the southeast indicating the general direction of groundwater flow through the aquifer while also displaying some variation from year to year. Despite some fluctuations, potentiometric surfaces for the Arbuckle-Simpson aquifer indicate that regional groundwater flow is toward the southeast, where streams and springs discharge at the aquifer boundary, which is consistent with the findings of Christenson and others (2011).

Water-level altitude measurements used for the 2022 potentiometric surface ranged from 861 to 1,163 ft above the North American Vertical Datum 1988 (NAVD 88), with a mean of 1,039 ft above NAVD 88 (USGS, 2024a). Water-level altitudes used for the 2023 potentiometric surface ranged from 885 to 1,220 ft above NAVD 88, with a mean of 1,050 ft above NAVD 88. Although the variance and difference in mean values from 2022 to 2023 are not substantial, the mapped potentiometric contours have differences. The largest differences in water-level altitudes between 2022 compared 2023 are in an area containing four wells east of Pennington Creek. The water-level altitudes were higher in 2023 than in 2022: 1,086.8, 1,081.9, 1,092.3, and 1,092.5 ft above NAVD 88 for 2022 and 1,079.9, 1,077.2, 1,169.4, and 1,219.7 ft above NAVD 88 for 2023, respectively. Even with other examples of localized differences between the two maps, general patterns remained similar in 2022 and 2023.

In Phase 1, three potentiometric-surface maps were prepared and analyzed. The measurements used to generate these maps were collected during August 7–16, 1995; June 2006; and September 2006 (Christenson and others, 2011). As expected, these maps feature a slope from a topographic high between Blue River and Pennington Creek just east of the confined part of the eastern Arbuckle-Simpson aquifer to the southeast. This slope to the southeast from a topographic high is common to all Arbuckle-Simpson potentiometric-surface maps. Although this overarching pattern holds true as illustrated for Phase 1 and Phase 2 maps (figs. 12–13), localized variations for different time periods were observed.

The three potentiometric-surface maps produced for Phase 1 exhibit contours that “V” upstream near Delaware Creek in the southeast part of the main aquifer area (figs. 18–20 in Christenson and others, 2011). This “V” of contours is still slightly indicated but is harder to detect in the Phase 2 potentiometric-surface maps (figs. 12, 13) compared to those from Phase 1. This difference in prominence of contours could be due to the higher density of measurements in this area in Phase 1 that allowed detection of smaller scale features. This interpretation of the effects of measurement density on the ability to detect smaller scale features can be extended to other differences between the potentiometric surface; approximately 90 groundwater-level measurements were used to create the 2006 map compared to 57 and 56 groundwater-level measurements used for the 2022 and 2023 maps, respectively.

Across the eastern part of the Arbuckle-Simpson aquifer, groundwater levels were slightly higher in altitude (nearer the ground surface) during Phase 2 compared to Phase 1. The variation in the potentiometric surfaces between Phase 1 and Phase 2 could be attributed to the time at which measurements were collected. For Phase 2, water levels were measured February 14–17, 2022, and February 13–16, 2023, during base-flow conditions. Collecting groundwater-level measurements during base-flow conditions reduces effects from evapotranspiration, peak water use, and precipitation. In contrast to the Phase 2 groundwater levels, Phase 1 groundwater levels were not measured during base-flow conditions.

Recharge

Even though aquifer recharge can be a result of many different processes or a combination of processes, the dominant recharge mechanism for the eastern part of the Arbuckle-Simpson aquifer is infiltration of precipitation through the soil. This precipitation recharge takes place when precipitation falls on the land surface and subsequently percolates into the unsaturated zone. Evapotranspiration is the combination of evaporation and transpiration, with evaporation being water that transpires from a surface to the atmosphere and transpiration being water taken up by plants and released as vapor into the atmosphere (Sharp, 2024). The amount of precipitation that successfully infiltrates into soil and passes through an unsaturated zone for recharge to the water table is dependent on many factors such as the amount of water stored in the unsaturated zone, the slope of the land surface, the composition of the rocks and soils that form the aquifer, the type of vegetation and general land use overlying the aquifer, and the intensity, season, and duration of precipitation. Owing to the substantial number of variables that affect recharge via infiltration from precipitation, estimates are difficult to quantify. Two methods were used in this study to estimate recharge: (1) a recession-curve displacement method, referred to as the RORA method (Rorabaugh, 1964; Rutledge, 1998) and implemented as the RORA program, was used to estimate recharge in a watershed upstream from a selected streamgage, and (2) the Soil-Water-Balance (SWB) code (Westenbroek and others, 2010) was used to model spatially distributed recharge rates across the aquifer.

Soil-Water-Balance Code

The SWB (Westenbroek and others, 2010) code was run to estimate the amount and spatial distribution of daily groundwater recharge to the eastern part of the Arbuckle-Simpson aquifer only for the years from 2019 through 2022 of the Phase 2 period because most of the groundwater well data collection in Phase 2 started in 2019 (table 8). At the time the SWB code was run for this report, the climate data for 2023 were not available yet. SWB is a modified Thornthwaite and Mather (1957) SWB method that uses gridded climate data and landscape characteristics to calculate recharge by using the following equation:

R

= (

P

+

S

+

R_i

) – (

Int

+

R_o

+

P_et

) –

ΔSm

,

(2)

where

R

is recharge, in inches per day;

P

is precipitation, in inches per day;

S

is snowmelt, in inches per day;

R_i

is surface runoff inflow, in inches per day;

Int

is plant interception, in inches per day;

R_o

is surface runoff outflow, in inches per day;

P_et

is potential evapotranspiration, in inches per day; and

ΔSm

is the change in soil moisture, in inches per day.

Data inputs that are used in the SWB code include precipitation, air temperature, soil-water storage capacity, hydrologic soil group, land-surface flow direction, and land-cover type. The user-specified grid for the eastern part of the Arbuckle-Simpson aquifer SWB code consisted of 2,180 columns by 1,831 rows of cells that were each 328 by 328 ft. Climate data inputs included daily grids of precipitation data and minimum and maximum air temperature during 2019–22 from the Daymet climate database (Thornton and others, 2024). Soil properties (soil-water storage capacity and hydrologic soil group) were obtained from the gridded Soil Survey Geographic database (U.S. Department of Agriculture, 2024). Land-cover types were obtained from the National Land Cover Database (Multi-Resolution Land Characteristics Consortium, 2024) and resampled to the SWB model grid resolution by using the most common land-cover type within each cell. Surface runoff was derived by using the D8 method (Greenlee, 1987) to calculate the land-surface gradient from a 10-meter DEM (USGS, 2015). Depressions were filled by using the ArcGIS Fill tool (Esri, 2024) after the DEM was resampled to the SWB model grid size. Filling depressions in the DEM ensures correct routing of surface runoff and eliminates isolated sink features that would result in unrealistically high amounts of recharge.

The Hargreaves and Samani (1985) method for calculating evapotranspiration was used for a reference latitude of 34.240000 to 34.740000 degrees. Land-cover types were used in conjunction with hydrologic soil groups to partition daily precipitation into plant interception (Int) and surface runoff (R_i and R_o) components and assign plant root-zone depths. The root-zone depths for grassland/herbaceous and pasture (the dominant land-cover types in the study area) varied with soil texture but ranged from about 0.8 to 1.5 ft. The maximum volume of water available in the root zone was calculated by multiplying the soil-water storage capacity by the root-zone depth. Changes in soil moisture (ΔSm) exceeding the soil-water storage capacity were assumed to be recharge (R) to the saturated zone. Smaller root-zone depths resulted in increased recharge and decreased evapotranspiration of water from the root zone whereas larger root-zone depths resulted in decreased recharge and increased evapotranspiration of water from the root zone. Recharge from irrigation was not simulated by SWB but was assumed to be negligible given the relatively small amount of irrigation groundwater use in the study area.

Recharge from precipitation was assumed to occur where the eastern part of the Arbuckle-Simpson aquifer is unconfined. The portion confined by the post-Simpson Pennsylvanian units was analyzed by using the SWB code but was not included in the mean recharge calculation for the study area (fig. 14). The mean annual SWB-estimated recharge for the eastern part of the Arbuckle-Simpson aquifer was 4.85 in/yr (or 13 percent of the 37.97 inches of mean annual precipitation) for the 2019–22 period (fig. 4A). The annual SWB-estimated recharge for the eastern part of the Arbuckle-Simpson aquifer was 7.2 inches for 2019, 5.9 inches for 2020, 2.6 inches for 2021, and 3.6 inches for 2022. Mean annual recharge estimated by using the RORA program for Phase 2 (2018–22) was 16.5 in/yr, which is greater than the mean annual recharge of 4.85 in/yr estimated by using SWB for Phase 2 (2019–22).

The SWB code cannot be used to simulate interactions between surface-water and groundwater features. In locations where the water table is beneath the bottom of the root zone, the SWB code can be used to produce reasonable annual or monthly values. The depth from the bottom of the root zone to the top of the water table is not considered in the estimation of recharge because there may be appreciable travel time through the unsaturated zone. Simulating the unsaturated zone was outside the scope of this project. Using the SWB code in areas with wetlands, springs, and lakes where the water table is close to the land surface produced unrealistic annual or monthly values because there is no provision for recharge rejection in the form of saturation excess (other than a maximum recharge rate that can be specified for a land use and soil type). Because the eastern part of the Arbuckle-Simpson aquifer is a karst aquifer, recharge estimates obtained by using the SWB code were compared with estimates obtained by using other recharge methods (such as the RORA program). SWB code can nonetheless be used to create reliable spatially gridded recharge estimates for the study area with spatial variations in recharge across different surficial features of the eastern part of the Arbuckle-Simpson aquifer. The alluvial sediments and soils near streams have higher recharge rates than the surrounding soil types overlying the Arbuckle and Simpson Groups that make up the eastern part of the Arbuckle-Simpson aquifer (fig. 14).

Groundwater Withdrawals

The amount of groundwater that can be appropriated in Oklahoma is tied to the amount of land the applicant either owns or leases and the equal proportionate share determined by the OWRB for the underlying aquifer (Oklahoma State Legislature, 2023b). These lands put toward the application are referred to as “dedicated lands” (fig. 1). Groundwater-use data are self-reported by the permitted groundwater user annually to the OWRB. Nine different beneficial groundwater-use types are regulated by the OWRB for groundwater permits: (1) irrigation, (2) public supply, (3) industrial, (4) power, (5) mining, (6) commercial, (7) recreation, fish, and wildlife, (8) agricultural, and (9) other (fig. 1). Power and commercial groundwater-use type was not included in this study’s results because of no reported use in that category. The amounts of groundwater used for recreation, fish, and wildlife, agriculture, and other have been combined into one category because of their relatively low combined use amount compared to the other uses. Public supply was the most reported beneficial groundwater-use type during 1967–2020 for the eastern part of the Arbuckle-Simpson aquifer, with the aquifer providing water for municipalities within the study area (fig. 15A; table 11; Mashburn and others, 2025). The mean annual reported groundwater use during 1967–2020 was 3,251 acre-feet.

Permitted amounts per groundwater user were summed for each year (fig. 15B). Some permitted groundwater users did not report use each year, which could be because of not using their permitted allotment or they were not compliant with reporting requirements. For example, the City of Ada reported groundwater use of 4,179 and 5,094 acre-feet in 2011 and 2012, respectively, with little (less than 1,100 acre-feet) to no groundwater use reported during 2013–20. As described in Christenson and others (2011, p. 51),

Consumptive Water Use at Producing Mines

Additional rules related to the withdrawal, use, and disposal of groundwater associated with producing mines provide a framework to protect groundwater in the Arbuckle-Simpson aquifer (Oklahoma State Legislature, 2023e). Based on these rules, the OWRB gathers data provided by the mining companies of monitoring and periodic reporting of groundwater disposition. Mines are required by the OWRB to submit either a management plan to monitor and quarterly and annual reports of water volumes. These management plans and reports include a water budget for the mine that describes anticipated water moving into and out of the mine site as stream and groundwater inflows and outflows (includes augmentation outflows to streams and the groundwater system), pit water, precipitation runoff, and evaporation. These water volumes are used by the OWRB to estimate the consumptive use of pit water.

Water volume data from these management plans and reports were compiled for 2019–22 from the OWRB website (OWRB, 2024b) in February 2024. Each mine is required to report these data once every quarter, as well as a total annual water volume. Each mine is required to report the amount of precipitation that falls onto the land and flows over the land surface and subsequently drains into mine pits, as well as nonconsumptive water losses (including pit water returned to the land surface from which surface runoff flows into a mine pit). Some of the mines, are considered de minimis, that coincide with the location of the eastern part of the Arbuckle-Simpson aquifer have an exemption from having to report the measured quarterly or annual water volumes, but they are still required to submit a plan of anticipated volumes. De minimis is the taking, use or disposal of pit water in an amount less than five acre-feet per year (Oklahoma State Legislature, 2023b). For the purposes of this study, water volumes provided per mine in either a management plan or a quarterly or annual report were included in the estimated consumptive use. The consumptive use estimates do not distinguish between groundwater and surface water. Some reports appeared to be incomplete; however, it was beyond the scope of this report to resolve any apparent discrepancies in reported water volumes.

Consumptive use water volumes were compiled from annual plans and reports for the 2019–22 period for eight mines (table 12). Of the 11 mines that submitted reports to the OWRB, 3 of them indicated that they had no water volume to report (OWRB, 2024b). Three of the remaining eight mines that submitted reports coincide with the surficial extent of the eastern part of the Arbuckle-Simpson aquifer (table 12). A negative consumptive water-use value indicates that there was no consumption of water by the mine and water was returned to the natural system through a variety of processes that exceeded the total volume of water withdrawn from the producing mine pit (mine pit pumping), including natural evaporation from the pit, water evaporated during the drying of the mined material, water in the mined material transported from the mine, volume of water used for beneficial uses, pit water returned to a groundwater basin (groundwater augmentation), pit water returned to a definite stream during flow conditions that are less than or equal to 50-percent exceedance (stream augmentation) (fig. 7), and other losses such as pit groundwater returned to the land surface via runoff. The 50-percent exceedance is the median of mean daily flows for the period of record analyzed (Oklahoma State Legislature, 2023b). The USGS streamflow data are used by the OWRB to update the calculation of the 50-percent exceedance annually. Mines can offset the water pumped from the mine if stream augmentation occurs when the streamflow is less than or equal to the 50-percent exceedance. The sum of the water withdrawn from mine pits, sum of stream augmentation, and sum of groundwater augmentation are also reported for each year for the eight mines for which compiled water-use data were available (table 12).

Water is discharged into streams from point sources by mines that have a National Pollutant Discharge Elimination System (NPDES) permit (EPA, 2024b). Eight mines located within the surficial extent of the eastern part of the Arbuckle-Simpson aquifer had an NPDES permit. Four of the eight mines reported discharge monitoring data for outflow discharges during 2019–23 (EPA, 2024b). The reported NPDES monthly mean of daily discharges ranged from 1.2 to 62 acre-feet per day during 2019–23. The reported NPDES monthly maximum of daily discharges ranged from 1.2 to 134 acre-feet per day during 2019–23.