This report describes a study of the hydrogeology and simulation of groundwater flow for the Canadian River alluvial aquifer in western and central Oklahoma conducted by the U.S. Geological Survey in cooperation with the Oklahoma Water Resources Board. The report (1) quantifies the groundwater resources of the Canadian River alluvial aquifer by developing a conceptual model, (2) summarizes the general water quality of the Canadian River alluvial aquifer groundwater by using data collected during August and September 2013, (3) evaluates the effects of estimated equal proportionate share (EPS) on aquifer storage and streamflow for time periods of 20, 40, and 50 years into the future by using numerical groundwater-flow models, and (4) evaluates the effects of present-day groundwater pumping over a 50-year period and sustained hypothetical drought conditions over a 10-year period on stream base flow and groundwater in storage by using numerical flow models. The Canadian River alluvial aquifer is a Quaternary-age alluvial and terrace unit consisting of beds of clay, silt, sand, and fine gravel sediments unconformably overlying Tertiary-, Permian-, and Pennsylvanian-age sedimentary rocks. For groundwater-flow modeling purposes, the Canadian River was divided into Reach I, extending from the Texas border to the Canadian River at the Bridgeport, Okla., streamgage (07228500), and Reach II, extending downstream from the Canadian River at the Bridgeport, Okla., streamgage (07228500), to the confluence of the river with Eufaula Lake. The Canadian River alluvial aquifer spans multiple climate divisions, ranging from semiarid in the west to humid subtropical in the east. The average annual precipitation in the study area from 1896 to 2014 was 34.4 inches per year (in/yr).

A hydrogeologic framework of the Canadian River alluvial aquifer was developed that includes the areal and vertical extent of the aquifer and the distribution, texture variability, and hydraulic properties of aquifer materials. The aquifer areal extent ranged from less than 0.2 to
8.5 miles wide. The maximum aquifer thickness was 120 feet (ft), and the average aquifer thickness was 50 ft. Average horizontal hydraulic conductivity for the Canadian River alluvial aquifer was calculated to be 39 feet per day, and the maximum horizontal hydraulic conductivity was calculated to be 100 feet per day.

Recharge rates to the Canadian River alluvial aquifer were estimated by using a soil-water-balance code to estimate the spatial distribution of groundwater recharge and a water-table fluctuation method to estimate localized recharge rates. By using daily precipitation and temperature data from 39 climate stations, recharge was estimated to average 3.4 in/yr, which corresponds to 8.7 percent of precipitation as recharge for the Canadian River alluvial aquifer from 1981 to 2013. The water-table fluctuation method was used at one site where continuous water-level observation data were available to estimate the percentage of precipitation that becomes groundwater recharge. Estimated annual recharge at that site was 9.7 in/yr during 2014.

Groundwater flow in the Canadian River alluvial aquifer was identified and quantified by a conceptual flow model for the period 1981–2013. Inflows to the Canadian River alluvial aquifer include recharge to the water table from precipitation, lateral flow from the surrounding bedrock, and flow from the Canadian River, whereas outflows include flow to the Canadian River (base-flow gain), evapotranspiration, and groundwater use. Total annual recharge inflows estimated by the soil-water-balance code were multiplied by the area of each reach and then averaged over the simulated period to produce an annual average of 28,919 acre-feet per year (acre-ft/yr) for Reach I and 82,006 acre-ft/yr for Reach II. Stream base flow to the Canadian River was estimated to be the largest outflow of groundwater from the aquifer, measured at four streamgages, along with evapotranspiration and groundwater use, which were relatively minor discharge components.

Objectives for the numerical groundwater-flow models included simulating groundwater flow in the Canadian River alluvial aquifer from 1981 to 2013 to address groundwater use and drought scenarios, including calculation of the EPS pumping rates. The EPS for the alluvial and terrace aquifers is defined by the Oklahoma Water Resources Board as the amount of fresh water that each landowner is allowed per year per acre of owned land to maintain a saturated thickness of at least 5 ft in at least 50 percent of the overlying land of the groundwater basin for a minimum of 20 years.

The groundwater-flow models were calibrated to water-table altitude observations, streamgage base flows, and base-flow gain to the Canadian River. The Reach I water-table altitude observation root-mean-square error was 6.1 ft, and 75 percent of residuals were within ±6.7 ft of observed measurements. The average simulated stream base-flow residual at the Bridgeport streamgage (07228500) was 8.8 cubic feet per second (ft³/s), and 75 percent of residuals were within ±30 ft³/s of observed measurements. Simulated base-flow gain in Reach I was 8.8 ft³/s lower than estimated base-flow gain. The Reach II water-table altitude observation root-mean-square error was 4 ft, and 75 percent of residuals were within ±4.3 ft of the observations. The average simulated stream base-flow residual in Reach II was between 35 and 132 ft³/s. The average simulated base-flow gain residual in Reach II was between 11.3 and 61.1 ft³/s.

Several future predictive scenarios were run, including estimating the EPS pumping rate for 20-, 40-, and 50-year life of basin scenarios, determining the effects of current groundwater use over a 50-year period into the future, and evaluating the effects of a sustained drought on water availability for both reaches. The EPS pumping rate was determined to be 1.35 acre-feet per acre per year ([acre-ft/acre]/yr) in Reach I and 3.08 (acre-ft/acre)/yr in Reach II for a 20-year period. For the 40- and 50-year periods, the EPS rate was determined to be
1.34 (acre-ft/acre)/yr in Reach I and 3.08 (acre-ft/acre)/yr in Reach II. Storage changes decreased in tandem with simulated groundwater pumping and were minimal after the first 15 simulated years for Reach I and the first 8 simulated years for Reach II.

Groundwater pumping at year 2013 rates for a period of 50 years resulted in a 0.2-percent decrease in groundwater-storage volumes in Reach I and a 0.6-percent decrease in the groundwater-storage volumes in Reach II. The small changes in storage are due to groundwater use by pumping, which composes a small percentage of the total groundwater-flow model budgets for Reaches I and II.

A sustained drought scenario was used to evaluate the effects of a hypothetical 10-year drought on water availability. A 10-year period was chosen where the effects of drought conditions would be simulated by decreasing recharge by 75 percent. In Reach I, average simulated stream base flow at the Bridgeport streamgage (07228500) decreased by 58 percent during the hypothetical 10-year drought compared to average simulated stream base flow during the nondrought period. In Reach II, average simulated stream base flows at the Purcell streamgage (07229200) and Calvin streamgage (07231500) decreased by 64 percent and 54 percent, respectively. In Reach I, the groundwater-storage drought scenario resulted in a storage decline of 30 thousand acre-feet, or an average decline in the water table of
1.2 ft. In Reach II, the groundwater-storage drought scenario resulted in a storage decline of 71 thousand acre-feet, or an average decline in the water table of 2.0 ft.