Purpose and Scope

The purpose of this report is to describe the hydrogeology and conceptualize groundwater flow for the part of the Red River alluvial aquifer contained in the alluvium and terrace deposits east of Lake Texoma in southeastern Oklahoma to the north-south Oklahoma-Arkansas State line. This report presents a hydrogeologic framework of the eastern part of the alluvial aquifer and documents the development of a conceptual groundwater-flow model of the aquifer during the study period 1980–2022. The part of the Red River alluvial aquifer in Oklahoma described herein is referred to as the “eastern part of the Red River alluvial aquifer.” Although the focus herein is on this part of the aquifer within Oklahoma, the study area also includes parts of the alluvium and terrace containing the Red River alluvial aquifer in Texas and Arkansas (fig. 1). Although the study areas differ, the organization and the wording of sections in this report are modified from those in Smith and others (2017, 2021).

Description of Study Area

The eastern part of the Red River alluvial aquifer in southeastern Oklahoma consists of approximately 401,280 acres of Quaternary alluvium and terrace deposits associated with the Red River and its major tributaries. The Red River alluvial aquifer is an expansive aquifer along the southern border of Oklahoma, although the geographic scope of this report is the alluvium and terrace deposits containing the Red River alluvial aquifer in Oklahoma between Lake Texoma and the Arkansas State line (fig. 1). Some analyses in this report include the alluvial and terrace deposits in Texas and Arkansas to address hydrologic influence from surrounding areas. Similar to the rest of the Red River alluvial aquifer, the eastern part of this aquifer is hydrologically connected to the Red River and its major tributaries and is composed of unconsolidated terrace and alluvial deposits associated with the Red River and selected tributaries (fig. 1).

Land Use

Land-use data for the study area in southeastern Oklahoma, northeastern Texas, and southwestern Arkansas were obtained from the CropScape database, which includes land-cover characteristics compiled at 30-meter (m) resolution for 2022 (figs. 2, 3; National Agricultural Statistics Service, 2024; U.S. Department of Agriculture [USDA], 2023). Land cover overlying the eastern part of the Red River alluvial aquifer is primarily cropland (81.6 percent), which is concentrated near the southern extent of the aquifer (figs. 2, 3). The remainder of the land-cover types are forest or shrubland (13.4 percent); developed (3.2 percent); and other, which includes water, wetlands, or barren land (1.9 percent). Predominant crops contributing to the total percentage of cropland in the study area are winter wheat (54.9 percent), cotton (17.7 percent), and hay or alfalfa (9.4 percent). Sorghum, corn, and soybeans account for a combined 8.4 percent of the total cropland. Fallow or idle cropland accounted for 9.5 percent of the cropland. The percentages for crop-cover types are long-term mean values and are subject to change in any given year because of seasonal, economic, and hydrologic factors. Percentages may not sum to 100.0 percent because of rounding.

Long-Term Climate Patterns

The climate of southeastern Oklahoma is characterized as humid subtropical (Kottek and others, 2006). Daily climate data (mean, maximum, and minimum temperature, and mean precipitation) were compiled from nine selected climate stations near or within the study area (fig. 1, table 1) operated as part of the Oklahoma Mesonet network (Oklahoma Mesonet, 2023) or the National Centers for Environmental Information (NCEI, 2023). Climate data were used to assess patterns in long-term (1916–2023) annual and monthly temperature and precipitation for the study area. A locally weighted scatterplot smoothing (LOWESS) curve (Cleveland, 1979) was applied to the annual data to determine periods of above- or below-mean precipitation or temperature (fig. 4).

Mean annual precipitation in the study area for the 1916–2023 period of record was 46.6 in., and the mean annual temperature was 63.3 degrees Fahrenheit (°F). Compared to those values, the mean annual precipitation was 1.3 inches (in.) higher, and the mean annual temperature was 0.3 °F lower for the 1980–2022 study period. Within the study area, precipitation data display temporal variations. Minimum annual precipitation was 27.9 in. during 2005, and maximum annual precipitation was 67.2 in. during 2015 (fig. 4A). Temperature is also variable over the period of record; the minimum annual mean temperature was 59.5 °F in 1926, and the maximum annual mean temperature was 67.0 °F in 1921 (fig. 4B).

Monthly precipitation is typically greatest in April and May, and steadily increases from January to May. Monthly precipitation is lowest in January, and annual snowfall in the study area is the lowest in the State (Oklahoma Climatological Survey, 2024). The mean monthly precipitation for the 1980–2022 study period was greatest (6.1 in.) in May and least (2.9 in.) in January (fig. 5A). The mean monthly temperature during 1980–2022 was highest (81.8 °F) in July and lowest (42.2 °F) in January (fig. 5B). Patterns in mean monthly precipitation and temperature were similar over the combined period of record for both precipitation and temperature (1916–2023) and during the study period (1980–2022).

Streamflow and Base-Flow Patterns

Daily streamflow data were recorded at selected USGS streamgages in the study area (fig. 1) and summarized for the 1980–2022 study period (table 2). Streamflow is primarily the sum of runoff and base flow; precipitation that falls directly on a stream is a small, nearly negligible component of streamflow (Barlow and Leake, 2012). Runoff refers to surface runoff (overland flow) that drains into a stream channel and subsurface stormflow (interflow) that originates from surface runoff. Base flow is the streamflow component supplied by groundwater discharge to streams (Barlow and Leake, 2012).

For this report, streamflow-hydrograph data obtained from the USGS National Water Information System (NWIS; USGS, 2024) were separated into runoff and base-flow components by using the Base-Flow Index (BFI) code (Wahl and Wahl, 1995) in the USGS Groundwater Toolbox (Barlow and others, 2015). The BFI code uses the minimum streamflow in a moving n-day window as a basis for hydrograph separation, where n is the user-defined number of days. Turning points are then determined by comparing minimums to adjacent minimums on the base-flow hydrograph. If 90 percent of a given minimum is less than the preceding and succeeding minimums, then that minimum is considered a turning point. Turning points can then be used to linearly interpolate base flow by connecting adjacent values with a straight line (Moix and Galloway, 2005). With these data, the BFI (or the percentage of base flow per streamflow) can be calculated. Multiple n-day bins were tested by plotting mean BFI against different n-day values. For consistency, a 5-day window and an f-statistic of 0.9 were used for all streamgages in this report.

Three USGS streamgages in the study area were used for BFI analysis: 07332500 Blue River near Blue, Okla. (map identifier S-02) (hereinafter referred to as the “Blue River streamgage”); 07335500 Red River at Arthur City, Tex. (map identifier S-04) (hereinafter referred to as the “Arthur City streamgage”); and 07337000 Red River at Index, Ark. (map identifier S-06) (hereinafter referred to as the “Index streamgage”). Data from the Blue River, Arthur City, and Index streamgages were analyzed for the period 1937–2022 to depict monthly base flow, monthly streamflow, and the annual base-flow index with an associated 5-year moving average (fig. 6). Streamgage data for the Arthur City and Index streamgages indicate BFI values generally increased over the first part of the 1937–2022 period of record and were in a slightly downward or stable pattern from the mid-1990s through 2022. Data from the Blue River streamgage indicate relatively consistent BFI values over the 1937–2022 period of record. Mean annual BFI values for the Blue River, Arthur City, and Index streamgages over the 1980–2022 period were 32.1, 52.0, and 58.3 percent, respectively (table 2). The mean annual base flows for the Blue River, Arthur City, and Index streamgages for the 1980–2022 period were 103.4; 5,670; and 5,170 cubic feet per second (ft³/s), respectively. For the overall 1937–2022 period of record, there is a slight upward pattern in annual base-flow values for all three sites (fig. 6A–C).

Groundwater Use

The OWRB permits and regulates all groundwater use except for groundwater withdrawals less than 5 acre-ft/yr used for domestic purposes for irrigating less than 3 acres of land for growing gardens, orchards, or lawns in Oklahoma (Oklahoma Statutes §82–1020.1[2] [Oklahoma State Legislature, 2021a]; OWRB, 2012c; Oklahoma Statutes §82–1020.3 [Oklahoma State Legislature, 2021c]). Since 1980, permitted users have self-reported groundwater-use data annually to the OWRB. Groundwater permits issued by the OWRB are divided into nine categories: (1) irrigation; (2) public supply; (3) industrial; (4) power; (5) mining; (6) commercial; (7) recreation, fish, and wildlife (hereinafter referred to as “recreation”); (8) agricultural; and (9) other. Groundwater use reports submitted to the OWRB by 702 permitted users that withdraw water from the eastern part of the Red River alluvial aquifer were summarized for 1967–2022 (OWRB, 2024b).

Most of the reported groundwater withdrawn from the eastern part of the Red River alluvial aquifer was used for irrigation and accounted for a long-term mean of 1,440 acre-ft/yr, or approximately 84 percent of the total reported groundwater use of 1,720 acre-ft/yr for 1967–2022 (fig. 7A; table 3). Over the same period, a mean of 263 acre-ft/yr, or approximately 15 percent of total groundwater use, was for public supply and the approximately remaining 1 percent was for agriculture (slightly more than 0.9 percent) and a combination of other, industrial, and recreation (slightly less than 0.1 percent). For the purposes of this long-term (1967–2022) analysis, the categories other, industrial, and recreation were combined, owing to their comparatively small shares of total groundwater use in the focus area (slightly less than 0.1 percent of total groundwater use combined). Although the proportions of individual permit categories fluctuated throughout the 1967–2022 period, irrigation was the predominant use category for most years analyzed (fig. 7B). Total reported groundwater use data also fluctuated throughout the study period but indicated an upward pattern overall, especially during 2012–22 for which mean annual groundwater use was 2,235 acre-ft/year greater than the 1967–2022 long-term mean (fig. 7B). This upward pattern in groundwater use is further indicated by comparing mean annual water use during the 1980–2022 study period to the 1967–2022 period of record in which reported nondomestic groundwater use increased approximately 277 acre-ft/yr (table 4). Changes in groundwater use were also observed to coincide with changes in the amount of precipitation falling on the eastern part of the Red River alluvial aquifer. During the 2007–09 period, groundwater use was relatively low (fig. 7B), and precipitation exceeded the long-term mean, which lessened the demand for groundwater withdrawals for irrigation (fig. 4A; Oklahoma Mesonet, 2023; NCEI, 2023). The portion of groundwater use reported for public supply was greater than usual compared to irrigation for the 1982–99 period, which was also a period when annual precipitation exceeded the long-term mean annual precipitation, thereby decreasing groundwater use for irrigation. Total permitted water use increased over the period of record and has not exceeded the total permitted amount (fig. 7B).

Currently (2025), the OWRB permits active groundwater wells within the study area to withdraw water for various use categories (fig. 8). Across the eastern part of the Red River alluvial aquifer, the predominant use type for these wells is irrigation, with a large concentration near Lake Texoma (fig. 8). All reported annual groundwater use data are published in the accompanying data release (Gammill and others, 2025).

Quaternary Sedimentary Deposits and Cretaceous Bedrock Units

The Quaternary sedimentary deposits in the study area primarily consist of alluvium and terrace deposits formed by the Red River and its tributaries (fig. 9; Ryder, 1996; Stoeser and others, 2005). Alluvium deposits contain sand, silt, and clay (Frye and Leonard, 1963; Huffman and others, 1978). Frye and Leonard (1963) subdivided terrace deposits along the Red River in northern Texas by land-surface altitude, slight compositional and structural differences, and fossil assemblages. The terrace deposits consist of gravel, sand, silt, and clay, and differ slightly in composition and thickness across the study area (Huffman and others, 1978).

Units that make up the bedrock beneath the eastern part of the Red River alluvial aquifer consist of consolidated units distinct from the alluvium or terrace deposits. In the study area, sedimentary bedrock units deposited during the Cretaceous underlie the alluvium or terrace deposits composing the eastern part of the Red River alluvial aquifer. The Oklahoma Geological Survey described geology at the county scale in Bryan, Choctaw, and McCurtain Counties (Davis, 1960; Huffman and others, 1975, 1978), the three counties east of Lake Texoma that contain the eastern part of the Red River alluvial aquifer in Oklahoma (fig. 1).

The Oklahoma Geological Survey (Huffman and others, 1978) conducted a geology and mineral resources study in Bryan County, documenting that Lower Cretaceous bedrock units from the Trinity Group are exposed at the surface and overlain by extensive Quaternary terrace and alluvium deposits near streams. Lower Cretaceous bedrock units are divided into the Trinity, Fredericksburg, and Washita Groups. Within this area, the Trinity Group is represented in Bryan County by sandstone exposures; the Fredericksburg Group is represented by limestone exposures, sometimes fossiliferous and interbedded with fissile shale; and the Washita Group is represented by sandstone, limestone (sometimes fossiliferous and interbedded with shale), limestone-shale, shale, and marlstone with nodular limestone exposures (Huffman and others, 1978).

The Oklahoma Geological Survey (Huffman and others, 1975) described the geology and mineral resources in Choctaw County, noting that Mississippian or Cretaceous bedrock units are exposed at the surface in Choctaw County. In areas where Lower Cretaceous bedrock units (Trinity, Fredericksburg, and Washita Groups) have been eroded away (or partially eroded away in the case of the Washita Group) in the southeastern part of Choctaw County, the Upper Cretaceous Woodbine Formation unconformably overlies basal layers of the Washita Group. The Woodbine aquifer is considered a minor bedrock aquifer by the OWRB (fig. 10). As in Bryan County, Cretaceous bedrock units in Choctaw County are overlain by extensive Quaternary terrace and alluvium deposits.

The Oklahoma Geological Survey (Davis, 1960) investigated the geology and groundwater resources of southern McCurtain County. In McCurtain County, the Upper Cretaceous Tokio and Woodbine Formations underlie the units containing the eastern part of the Red River alluvial aquifer along most of the southern part of the county. In the southeastern part of McCurtain County along its eastern border, a small exposure of the undifferentiated Upper Cretaceous Ozan and Brownstown Formations underlie units containing the eastern part of the Red River alluvial aquifer. As in Bryan and Choctaw Counties, the Cretaceous bedrock units in McCurtain County are overlain by extensive Quaternary terrace and alluvium deposits.

Groundwater Levels

Hydrologic stressors, such as drought and flooding, can cause groundwater-level fluctuations on a spatial and temporal basis. Various hydrologic stressors affect groundwater-level fluctuations in an alluvial aquifer, including recharge, evapotranspiration, groundwater withdrawal, and streambed seepage (Freeze and Cherry, 1979). In 2021, continuous groundwater-level recorders were installed in six pre-existing wells completed in the eastern part of the Red River alluvial aquifer (wells Red01–Red06) (fig. 1; table 1; USGS, 2024). Seasonal groundwater-level patterns were similar for these six wells (fig. 11A–D). Groundwater levels typically decreased during summer and winter and increased during spring and fall. Precipitation data collected at nearby USGS streamgages were compared to groundwater levels from each USGS groundwater well. Groundwater levels measured at wells Red02, Red05, and Red06 displayed large changes in response to precipitation, whereas the changes in groundwater levels measured at wells Red03 and Red04 were muted following precipitation. The pattern of changes in groundwater levels recorded in wells Red03 and Red04 in response to groundwater withdrawals from nearby irrigation wells was characterized by rapid decreases in groundwater levels when irrigation withdrawals were occurring and slow recoveries after the withdrawals ceased. The groundwater-levels measured in well Red01 do not change appreciably as a result of changes in groundwater withdrawals for irrigation or in response to precipitation, but they do display seasonal fluctuations similar to those observed in the groundwater levels measured in wells Red02, Red05, and Red06.

Water Quality

Because of the considerable surface-water-groundwater interaction in alluvial aquifers (Winter, 1995), a combination of surface-water and groundwater samples were used to assess water quality in the eastern part of the Red River alluvial aquifer. The OWRB collected groundwater-quality samples at sites in the study area in August 2015 and again in June 2019 as part of their Groundwater Monitoring and Assessment Program (OWRB, 2018). In addition to the groundwater-quality data collected by the OWRB, surface-water-quality data were obtained from the NWIS database (USGS, 2024) from USGS sites in the study area sampled between October 1, 1959, and October 27, 2011. Water-quality data from 21 samples collected at 16 OWRB groundwater sites and 189 samples collected at 3 USGS streamgage sites (S-01, S-04, S-05; fig. 1; table 1) in the study area were used for analysis after omitting samples with incomplete data or samples with an ionic charge-balance error exceeding 20 percent after the major-ion concentrations were converted into milliequivalents (Hem, 1985). USGS surface-water samples used in this analysis were collected from the Red River.

Major cations and anions were examined using the Piper (1944) method (fig. 12). For this method, the selected major cations were calcium (Ca²⁺), magnesium (Mg²⁺), potassium (K⁺), and sodium (Na⁺), and the selected major anions were bicarbonate (HCO₃⁻), carbonate (C0₃²⁻), chloride (Cl⁻), fluoride (F⁻), and sulfate (SO₄²⁻). Samples lacking data for more than two major ions were not included in the analysis. Water-quality data were converted from milligrams per liter to milliequivalents per liter before the charge-balance error was calculated for each sample by using the following equation:

Surface-water quality samples tended to tightly cluster on the Piper diagrams and proportionally contained higher concentrations of magnesium, sodium, potassium, sulfate, and chloride, with dominant water types of Ca-Mg-SO₄-Cl and Na-SO₄-Cl. OWRB groundwater-quality samples were less clustered compared to surface-water samples. Groundwater-quality samples displayed higher concentrations of sodium, potassium, bicarbonate, carbonate, and chloride (fig. 12).

Dissolved-solids concentrations in the reported water-quality samples ranged from 40 to 1,537 mg/L, with a mean concentration of 893 mg/L and a median concentration of 977 mg/L (Gammill and others, 2025). The U.S. Environmental Protection Agency (2017) has established a secondary drinking-water standard of 500 mg/L for dissolved-solids concentrations, and the State of Oklahoma designates a domestic beneficial use for groundwater with dissolved-solids concentrations of less than 3,000 mg/L (OWRB, 2015). Dissolved-solids concentrations in groundwater-quality samples tended to be lower compared to those in surface-water samples. Results from OWRB groundwater-quality samples used for this analysis indicated dissolved-solids concentrations ranged from 40 to 1,064 mg/L, with a mean of 385 mg/L. In contrast, dissolved-solids concentrations in surface-water samples ranged from 106 to 1,537 mg/L, with a mean of 947 mg/L. Altogether, 179 of the 209 total number of samples used for Piper (1944) analysis exceeded the U.S. Environmental Protection Agency secondary drinking-water standard of 500 mg/L for dissolved-solids concentrations.

Textural and Hydraulic Properties

The distribution and variability of textural and hydraulic properties of aquifer materials were assumed to be the primary controls on groundwater flow in the eastern part of the Red River alluvial aquifer. Methods used to estimate the range of hydraulic conductivity values in the aquifer included hydraulic profiling in test holes, analysis of cored material, summary of lithologic logs, and a multiwell aquifer test.

Horizontal Hydraulic Conductivity Estimated From Test Holes and Cores

A direct-push Geoprobe hydraulic profiling tool (HPT) (Geoprobe Systems, 2015) was used to estimate horizontal hydraulic conductivity at nine test holes (Kh01–Kh09) across the eastern part of the Red River alluvial aquifer (fig. 1; table 1). To estimate horizontal hydraulic conductivity, water was injected at 0.05-ft depth intervals, and electrical conductivity, injection pressure (corrected for the hydrostatic pressure gradient when saturated), and injection rate were recorded. Horizontal hydraulic conductivity was calculated from data collected in each test hole at selected depth intervals by using the ratio of injection flow rate and injection pressure (Geoprobe Systems, 2015). This ratio method (McCall, 2010) is appropriate for estimating horizontal hydraulic conductivity where it is less than 150 feet per day (ft/d) in the saturated zone of an unconsolidated aquifer. All nine HPT sites were drilled to a depth of refusal, which was assumed to be the contact of the alluvium and terrace deposits with underlying bedrock units (Rogers and others, 2023).

The nine test holes were drilled to a range of 28.4 to 56.7 ft below land surface (table 5). At these test holes, the estimated depth to water ranged from 17.3 to 42.9 ft below land surface, and the mean horizontal hydraulic conductivity at each test hole ranged from 7.5 to 113 ft/d, with an overall mean of 62.2 ft/d (table 6). The distribution of discrete HPT horizontal hydraulic conductivity values is depicted in fig. 13A, B.

Table 5.    

Horizontal hydraulic conductivities calculated by using a Geoprobe hydraulic profiling tool at nine test holes in southeastern Oklahoma and northeastern Texas.

Table 5.    Horizontal hydraulic conductivities calculated by using a Geoprobe hydraulic profiling tool at nine test holes in southeastern Oklahoma and northeastern Texas.

Map identifier (fig. 1; table 1)	Total depth below land surface (feet)	Mean estimated horizontal hydraulic conductivity (feet per day)
Kh01	54.8	47.7
Kh02	31.3	34.8
Kh03	28.4	7.50
Kh04	43.2	25.3
Kh05	52.4	113
Kh06	49.2	50.2
Kh07	53.9	45.5
Kh08	56.7	52.5
Kh09	47.4	70.4

Table 6.    

Summary statistics of horizontal hydraulic conductivities obtained at nine test holes by using a Geoprobe hydraulic profiling tool (HPT) in southeastern Oklahoma and northeastern Texas (fig. 1; tables 1, 5).

Table 6.    Summary statistics of horizontal hydraulic conductivities obtained at nine test holes by using a Geoprobe hydraulic profiling tool (HPT) in southeastern Oklahoma and northeastern Texas (fig. 1; tables 1, 5).

Statistic	HPT estimated depth to water (feet below land surface)	Estimated horizontal hydraulic conductivity (feet per day)
Minimum	17.3	0.20
Mean	30.7	62.2
Median	34.3	40.5
Maximum	42.9	150

At one HPT test-hole site, Kh02 (fig. 1; table 1), a 27-foot core sediment sample was collected to check the accuracy of the HPT horizontal hydraulic conductivity measurements. Sediment cores were retrieved in 48-in.-long plastic tubes with a 2.25-in. diameter. The sediment core measured in the Kh02 test hole, like all other HPT measurements, was obtained by drilling from land surface to a depth of refusal, which was assumed to be the bedrock contact. Sediment cores were described in small depth increments by noting changes to grain size, sorting, and Munsell (1912) color; the data produced by this effort are included in the accompanying data release (Gammill and others, 2025). The lithologic descriptions of the sediment cores were grouped into one of six lithologic categories based on the dominant grain size: fine sand, medium sand, coarse sand, gravel, and clay. Lithologic descriptions obtained from the core were then correlated to the HPT values at the same depth to estimate horizontal hydraulic conductivity for each lithology (table 7). Mean horizontal hydraulic conductivity for the core collected at the Kh02 test hole was 29 ft/d (table 7).

Table 7.    

Horizontal hydraulic properties of lithologic categories calculated for a core collected at the Kh02 test hole in southeastern Oklahoma.

Table 7.    Horizontal hydraulic properties of lithologic categories calculated for a core collected at the Kh02 test hole in southeastern Oklahoma.

Lithologic description	Proportional amount of total core (percent)	Estimated horizontal hydraulic conductivity (feet per day)
		Minimum	Mean	Maximum
Fine sand	22.4	0	0	0
Medium sand	53.0	0.10	30	94
Coarse sand	0.70	0	0	0
Gravel	3.70	27	58	79
Topsoil	5.00	0	0	0
Clay	15.2	71	85	93
Total core	100	0	29	94

Hydraulic Properties Estimated From a Multiwell Aquifer Test

Multiwell aquifer tests are a more accurate method for estimating mean hydraulic conductivity and storage properties of aquifer materials compared to other methods such as deriving conductivity from grain size (Freeze and Cherry, 1979). A multiwell aquifer test was completed in November 2023 at an irrigation well in the eastern part of the Red River alluvial aquifer (well Red07; fig. 1; table 1). To conduct the test, well Red07 groundwater was withdrawn at an approximately constant rate of 316 gallons per minute for approximately 21 hours until the water levels measured in the designated observation well for the aquifer test (well Red04) stabilized (fig. 14A). The observation well Red04 was approximately 233 ft from the pumping well (well Red07). Pumping induced a maximum drawdown of 25.8 ft in the pumping well (Red07) (fig. 14A) and 1.33 ft in the observation well (Red04) (fig. 14C).

Water levels from the pumping and recovery period were matched to a curve defined by the Tartakovsky and Neuman (2007) method, which computes drawdown in an unconfined aquifer using the following equation:

$S=ST+SH+SU$

,

(3)

where

S

is the drawdown in an unconfined aquifer;

ST

is the drawdown for a fully penetrating well in a nonleaky confined aquifer;

SH

is the drawdown correction for partial penetration in a nonleaky confined aquifer; and

SU

is the drawdown correction for saturated-unsaturated flow in an unconfined aquifer.

Drawdown and recovery data obtained using the Tartakovsky and Neuman (2007) method were reported in either standard time or Agarwal equivalent time—an adjustment that transforms standard time into an equivalent time that is relative only to recovery time (Duffield, 2025). The Tartakovsky and Neuman (2007) method is most appropriate for evaluating groundwater flow in an unconfined aquifer, such as the eastern part of the Red River alluvial aquifer. The multiwell aquifer test was analyzed by using the AQTESOLV software package (Hydrosolve, Inc., 2011). Hydraulic properties presented in this report were collected from the observation well (Red04). Transmissivity, specific yield, and the storage coefficient were estimated by using the Tartakovsky and Neuman (2007) method with a user-specified saturated thickness value. A saturated thickness value of 116.5 ft was used for the observation well (Red04) and was computed from the saturated thickness map as discussed in the “Potentiometric Surface and Saturated Thickness” section herein. Transmissivity in the observation well was approximately 7,900 square feet per day (ft²/d). Geohydrologic-unit hydraulic conductivity was estimated at 67.49 ft/d. Specific yield (unitless) was estimated at 0.038, and the estimated storage coefficient (also unitless) was estimated at 0.002 (fig. 14).

Aquifer Extent

The spatial extent of the eastern part of the Red River alluvial aquifer was determined by using 1:250,000-scale geologic maps (Stoeser and others, 2005). The eastern part of this alluvial aquifer extends vertically from land surface to the depth of the geologic contact of the alluvium and terrace deposits with underlying bedrock units. The spatial extent of the aquifer is further defined by the eastern border of Lake Texoma, the Oklahoma-Texas State line, and the confluence of the Blue River with the Red River.

Where the eastern part of the Red River alluvial aquifer is present, the top of the Red River alluvial aquifer was defined as the land-surface altitude obtained from a 10-m (horizontal resolution) digital elevation model (DEM) (USGS, 2015) with filled depressions. The altitude of the base of the eastern part of the aquifer was contoured at a 20-ft interval from bedrock depths obtained from lithologic logs, well-completion reports, and test-hole data (OWRB, 2023; TWDB, 2023). The potentiometric surface of the aquifer was constructed using groundwater-level data and was used to determine saturated thickness for the aquifer by subtracting the base-of-aquifer altitude. The altitude of the base of the aquifer was calculated by subtracting the measured bedrock depth from the land-surface altitude. For consistency, the land-surface altitude was obtained from the 10-m DEM, even when the data source provided a land-surface altitude.

Lithologic logs obtained from the OWRB (2023) and the TWDB (2023) were used to construct the alluvium and bedrock surface contact of the Red River alluvial aquifer in records where the bedrock contact could be determined. Permian bedrock unit terms were used to identify red or consolidated materials such as “red bed,” “bedrock,” and “shale.” The bedrock surface was defined at the base of the aquifer by the presence of a red sand or shale on an impervious red clay layer adjoining an alluvial sand or gravel. The altitude of the base of the aquifer, determined from well logs, was contoured at a 20-ft interval. The altitude of the base of the eastern part of the Red River alluvial aquifer (fig. 15) ranged from 220 to 580 ft. The base of the aquifer dips downward from the aquifer’s border with Lake Texoma in the west eastward toward the Arkansas State line.

Potentiometric Surface and Saturated Thickness

A potentiometric surface is a theoretical topographic surface that displays the altitude at which water would have stood in tightly cased wells at a specified time across a given aquifer (Fetter, 2001). Potentiometric surface maps can be used to identify general groundwater-flow directions and delineate subsurface groundwater basins within an aquifer. Groundwater generally flows perpendicularly to contours from areas of high potentiometric altitudes to areas of low potentiometric altitudes (Freeze and Cherry, 1979).

A potentiometric surface map was constructed for the eastern part of the Red River alluvial aquifer by interpolating synoptic groundwater-level altitude data to create a continuous surface. Synoptic data were measured at wells constructed with steel or polyvinyl chloride casings with sand backfill surrounding the outside of the casings. To determine groundwater-level altitudes, the depth to water below land-surface datum was measured at each well by using a calibrated electric tape following methods described in Cunningham and Schalk (2011). The depth to water was subtracted from the land surface datum, referenced to the North American Vertical Datum of 1988 (NAVD 88), to compute the groundwater-level altitude at each well (Cunningham and Schalk, 2011). A stream is an expression of the groundwater-level altitude, so additional water-surface altitudes along streams were obtained from the 10-meter DEM and used as control points to create the potentiometric surface map (USGS, 2015). The groundwater-level altitudes, including the water-surface altitudes along streams, were interpolated by using inverse-distance-weighted interpolation methods in ArcGIS (Esri, 2024a). Contours were constructed from the interpolated surface and compared to altitude data from the DEM to ensure that potentiometric contours were below land surface.

During February 28–March 2, 2022, synoptic groundwater-level-altitude measurements were made at 44 wells across the Red River alluvial aquifer study area. Depth to the water table ranged from 1.36 to 65.85 ft below land surface datum. Additional groundwater-level altitude measurements were compiled from OWRB well-completion reports, which provided the interval depth at which water was first observed during the well drilling process (OWRB, 2023). Groundwater generally flowed in the same east-west direction as the flow of the Red River from a water-level altitude of approximately 680 ft near Lake Texoma in the west to 240 ft near the Arkansas State line in the east (fig. 16).

Saturated thickness was determined by subtracting the estimated altitude of the aquifer base—specifically, the contact between the base of alluvium that contains the aquifer and the top underlying bedrock—from the simulated groundwater-level altitude. Data from 211 lithologic logs were carefully checked to eliminate inaccurate point data, such as negative saturated thickness values that could affect saturated thickness interpolation, and then used to estimate saturated thickness (fig. 17). Saturated thickness was not calculated in parts of the study area where data were not available (fig. 17). Given the insufficient quantity of groundwater-level altitude measurements for parts of the aquifer within 1 mi of the Red River, some saturated thickness values were initially simulated as negative values. To eliminate negative saturated thickness values simulated within 1 mi of the Red River, a minimum of 20 ft of saturated thickness was required. This assumed minimum value is considered conservative based on the assumed 35–50 ft of saturated thickness of the similar sand-channel Washita alluvial aquifer (Hart, 1965). Bedrock altitudes determined from lithologic log point-data were used to create a topographic raster that in turn was used to create bedrock altitude contours. The bedrock-altitude contours were compared to potentiometric surface contours to ensure that bedrock contours were always lower in altitude than potentiometric contours. Simulated saturated thickness within the aquifer included a minimum thickness of 0 ft, a maximum thickness of 141.19 ft, and a mean thickness of 62.13 ft.

Hydrologic Boundaries

For the purposes of the conceptual model discussed in this report, hydrologic boundaries represent sources (inflows) and sinks (outflows) of water to and from the eastern part of the Red River alluvial aquifer. Some hydrologic boundaries can function as both sources of inflows and sinks for outflows and were classified as locations of “net inflows” or “net outflows” in this report, depending on which is the dominant flow component.

Recharge

For the purposes of this report, recharge is the infiltration of precipitation through the soil zone to the saturated zone. Infiltration occurs when precipitation falls on the land surface and percolates downward into the unsaturated zone (Freeze and Cherry, 1979). Many factors affect the amount of recharge to the aquifer, such as the amount of water stored in the unsaturated zone, the slope of the land surface, the composition of rocks and soils that contain the aquifer, the type of vegetation and general land use over the aquifer, and the intensity and duration of precipitation. Because many factors affect recharge, estimates are difficult to quantify. Two methods were used in this study to estimate recharge: the water-table fluctuation (WTF) method, used to estimate local recharge for the 2021–23 period, and application of the Soil-Water-Balance (SWB) code, used to model regional, spatially distributed recharge rates across the study area for 1980–2022 study period.

WTF Method

The WTF method (Healy and Cook, 2002) was used to estimate localized recharge for the eastern part of the Red River alluvial aquifer. The WTF method assumes that short-term level rises, lasting hours to days, in unconfined groundwater wells are attributable to precipitation causing recharge in the saturated zone of an aquifer. The WTF method is best applied to groundwater wells placed in locations with a shallow water table that display rapid responses to precipitation. The WTF method requires knowledge or estimation of specific yield, which can be variable across an aquifer. The WTF method was used to directly calculate recharge at specified groundwater wells using the following formula:

$R=S_y\text{* }\sum \frac{\Delta h}{\Delta t}$

,

(4)

where

R

is recharge, in inches per year;

S_y

is the specific yield of the aquifer (dimensionless);

Δh

is the change in water-level altitude, in inches; and

Δt

is the change in time, in years.

Water-level hydrographs from four of the six USGS continuous groundwater-level recording wells in the study area were used for WTF analysis because their hydrographs and locations were appropriate for the requirements and assumptions of the WTF method (Healy and Cook, 2002); the four wells used in the analysis were wells Red02, Red03, Red05, and Red06 (fig. 11A–D; table 9; USGS, 2024). Well Red01 was excluded from WTF analysis because of inadequate signals in water-level fluctuation, whereas well Red04 was excluded because the groundwater withdrawals at this well were much larger than the groundwater withdrawals at any of the other wells. The groundwater withdrawals at wells Red02, Red03, Red05, and Red06 were analyzed for 2022 and 2023. A specific yield ($S_y$) value of 0.038 was estimated by using the Tartakovsky and Neuman (2007) method with the data collected from observation well Red04 during the multiwell aquifer test and was used for WTF analysis. Daily precipitation data for the period 1980–2023 were obtained from climate sites closest to each USGS continuous groundwater-level recording well (fig. 1; table 1; Oklahoma Mesonet, 2023; NCEI, 2023).

Table 9.    

Summary of recharge estimates using the water-table fluctuation method for the Red River alluvial aquifer in southeastern Oklahoma and northeastern Texas, 2022–23.

[NCEI, National Centers for Environmental Information; Dates in mm/dd/yyyy format. Continuous water-level recorder wells Red01 and Red04 were not suitable for analysis with the water-table-fluctuation method. Estimated specific yield values are provided in E notation. %, percent; --, not applicable]

Table 9.    Summary of recharge estimates using the water-table fluctuation method for the Red River alluvial aquifer in southeastern Oklahoma and northeastern Texas, 2022–23.

Descriptor	U.S. Geological Survey continuous water-level-recorder well (fig. 1; table 1)
	Red02	Red03	Red05	Red06
Mean annual precipitation during 1980–2022, in inches, from nearby NCEI climate stations	46.2	46.2	44.1	47.6
Applicable NCEI or Oklahoma Mesonet climate stations	C-04, C-07	C-04, C-07	C-01, C-02	C-06, C-09
Estimated specific yield	3.8E-02	3.8E-02	3.8E-02	3.8E-02
Year 1, date range	01/01/2022–12/31/2022	01/01/2022–12/31/2022	01/01/2022–12/31/2022	01/01/2022–12/31/2022
Station annual precipitation, in inches	43.0	43.0	37.0	48.1
Sum of water-level rises, in feet	22.5	27.5	34.5	35.4
Recharge, in inches per year	10.3	12.5	15.7	16.2
Recharge, expressed as the percentage of annual precipitation	23.8%	29.1%	42.4%	33.6%
Recharge, in inches per year, normalized to mean annual precipitation, 1980–2022	11.0	13.5	18.7	16.0
Year 2, date range	01/01/2023–12/31/2023	01/01/2023–12/31/2023	01/01/2023–12/31/2023	01/01/2023–12/31/2023
Station annual precipitation, in inches	42.0	42.0	45.0	58.3
Sum of water-level rises, in feet	36.6	31.8	35.6	36.2
Recharge, in inches per year	16.7	14.5	16.2	16.5
Recharge, expressed as the percentage of annual precipitation	39.6%	34.5%	36.1%	28.3%
Recharge, in inches per year, normalized to mean annual precipitation, 1980–2022	18.3	16.0	15.9	13.5
Mean annual recharge 2022–2023, in feet per year, normalized to mean annual precipitation, 1991–2020	--	--	--	1.33

WTF-estimated mean annual recharge for the 2022–23 period was 16.0 inches per year (in/yr) (approximately 35.1 percent of mean annual precipitation for the period 1980–2022) (table 9; Oklahoma Mesonet, 2023; NCEI, 2023). The 1.33-ft/yr recharge rate multiplied by the 401,280-acre extent of the eastern part of the Red River alluvial aquifer in Oklahoma is equivalent to an estimated inflow volume of approximately 534,000 acre-feet per year (acre-ft/yr).

SWB Code

Spatially distributed groundwater recharge for a selected grid area within the study area (fig. 19) was estimated by using the SWB code (SWB version 1.2.511; Westenbroek and others, 2010). The SWB code uses a gridded data structure to compute the amount of daily infiltration resulting from precipitation, accounting for losses, that exceeds the storage capacity of the plant root zone (the depth below ground surface to which plant roots extend) and evapotranspiration demand from vegetation. Data required by the SWB code include climatological characteristics (precipitation and temperature) and landscape characteristics (soil-water storage capacity, hydrologic soil group, land-cover type, and land-surface gradient). The input data files and output recharge data files discussed in this report are included in the accompanying data release (Gammill and others, 2025). The SWB code uses the following equation to estimate recharge (modified from Westenbroek and others, 2010):

$R=\left(P+S+R_i\right)-\left(Int+R_0+P_{et}\right)-\Delta Sm$

,

(5)

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, in inches per day;

Int

is plant interception, in inches per day;

R₀

is surface runoff outflow, in inches per day;

P_et

is reference evapotranspiration, in inches per day; and

$\Delta Sm$

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

The SWB code estimates recharge (R) over a user-defined grid by using a modified Thornthwaite-Mather method (Thornthwaite and Mather, 1957). To estimate reference evapotranspiration (P_et) in the study area, the Hargreaves and Samani (1985) method for a reference latitude of 33.4–34.2 degrees was applied by using the SWB code. Land-cover data from the National Land Cover Database (Multi-Resolution Land Characteristics Consortium, 2023) were resampled to the SWB model grid resolution using the most common land-cover type within each cell. Land-cover data and hydrologic soil group data obtained from the Gridded Soil Survey Geographic database (USDA, 2021) were used to separate daily precipitation (P) into plant interception (Int) and surface runoff (R_i and R₀). To avoid unrealistic overestimations of recharge related to surface runoff, the D8 method (Greenlee, 1987) was used to calculate land-surface gradient from a 10-meter resolution DEM (USGS, 2015) that was resampled to the resolution of the SWB grid. Depressions, or sinks, in the DEM were filled using the ArcGIS Fill tool to ensure correct routing of surface runoff (Esri, 2024b). Daily precipitation, minimum temperature, and maximum temperature grids for the 1980–2022 SWB model period were obtained from the Daymet database (version 4; Thornton and others, 2020). Default plant root-zone depths for the SWB code, which are modeled from permeable glacial deposits in Wisconsin, were scaled to 25 percent for this analysis. Smaller root-zone depths resulted in increased recharge and decreased evapotranspiration of water from the plant root zone. SWB-estimated mean annual recharge within the eastern part of the Red River alluvial aquifer extent was 8.62 in/yr, or 17.98 percent of the mean annual precipitation over the same period (47.94 in/yr). The SWB-estimated mean annual recharge rate of 8.62 in/yr was approximately half of the WTF-estimated mean annual recharge rate of 16.0 in/yr for the 1980–2022 study period (fig. 19).

SWB-estimated recharge rates were summarized annually and monthly in conjunction with precipitation data to calculate recharge efficiency, which is the ratio of recharge to precipitation. Annual SWB-estimated recharge rates for southeastern Oklahoma ranged from 2.82 to 16.60 in/yr for the period 1980–2022, corresponding to years of below-mean and above-mean precipitation, respectively (figs. 4A, 20A). During the 1980–2022 study period, monthly recharge efficiency was equal to or above 20 percent from November to March, when evapotranspiration tends to be at a minimum because of decreased temperatures and precipitation (figs. 5, 20B). Monthly recharge efficiency did not exceed 15 percent for the warmer, wetter months of April through September, when evapotranspiration is generally higher compared to the mostly cooler months of October through March (figs. 5, 20B).

SWB-estimated recharge within the aquifer extent was used to approximate recharge for the eastern part of the Red River alluvial aquifer conceptual model because the SWB code accounts for various inputs over the entire aquifer, as opposed to the localized estimates gathered by the WTF method. Recharge was estimated as an inflow of 288,250 acre-ft/yr or 98.7 percent of the estimated inflows of the conceptual model water budget (fig. 18; table 8).

Saturated-Zone Evapotranspiration

Evapotranspiration is the process by which water is transferred to the atmosphere directly through evaporation and indirectly through plant transpiration. Most evapotranspiration occurs at the land surface where precipitation pools as surface water or where it infiltrates the soil unsaturated zone and becomes available to plant root zones; most precipitation does not reach the aquifer saturated zone (Lubczynski, 2009). Saturated-zone evapotranspiration mostly occurs in areas of the aquifer where the water table is close to the land surface and the saturated zone intersects the plant root zones, commonly in lower altitude areas such as near streams and wetlands (Lubczynski, 2009). Saturated-zone evapotranspiration was an essential consideration for the conceptual-model water budget, as approximately 58 percent, or 232,926 acres, of the eastern part of the Red River alluvial aquifer was classified as wetlands according to the U.S. Fish and Wildlife Service’s National Wetlands Inventory (fig. 21; table 8; U.S. Fish and Wildlife Service, 2023). Wetlands are land areas that are frequently saturated or flooded and are heavily concentrated along the Red River within the eastern part of the Red River alluvial aquifer. The National Wetlands Inventory’s classification of wetlands is different than that of the National Agricultural Statistics Service’s classification of wetlands, as the National Agricultural Statistics Service only classified 1.9 percent of the eastern part of the Red River alluvial aquifer as other, which includes wetlands, water, and barren land (figs. 2–3; National Agricultural Statistics Service, 2024; USDA, 2023). The National Wetlands Inventory is built upon a biological definition of wetlands, whereas the National Agricultural Statistics Service is focused on overlying vegetation used for agricultural purposes (U.S. Fish and Wildlife Service, 2024; National Agricultural Statistics Service, 2024).

Conceptual-Model Water Budget

The conceptual-model water budget (fig. 18; table 8) summarizes mean annual inflows to, and outflows from, the eastern part of the Red River alluvial aquifer during 1980–2022. Recharge, estimated from the SWB method, accounts for 98.7 percent of the conceptual-model inflows to the eastern part of the Red River alluvial aquifer. Saturated-zone evapotranspiration accounts for 11.9 percent, and net streambed seepage accounts for 87.4 percent of the outflows for the conceptual model (table 8). The contributions to outflows from saturated zone evaporation and net streambed seepage are essentially the inverse of those for other alluvial aquifers in Oklahoma (Ryter and Correll, 2016; Ellis and others, 2017, 2020; Smith and others, 2017, 2021; Rogers and others, 2023), likely because the part of the study area overlain by wetlands for the eastern part of the Red River alluvial aquifer was much larger compared to most alluvial aquifers in Oklahoma. In the eastern part of the Red River alluvial aquifer, a total of 232,926 acres, or 58 percent of the study area, is covered by wetlands (fig. 21; table 8), which could cause the higher percentage of saturated-zone evapotranspiration as seen in the conceptual model. Comparatively higher saturated-zone evapotranspiration could also be due to the comparatively warmer and wetter environment of the eastern part of the Red River alluvial aquifer (Kottek and others, 2006), because warmer temperatures allow for more atmospheric evapotranspiration storage and more frequent or abundant precipitation may allow greater availability of groundwater to the plant root zone. Net changes in groundwater storage (1.3 percent), and well withdrawals (0.7 percent) each accounted for less than 5 percent of conceptual water-budget inflows or outflows. The net change in groundwater storage of the eastern part of the Red River alluvial aquifer was assumed to be a small component of the conceptual model and was used to balance the water budget.