Simulation of groundwater flow and analysis of projected water use for the Rush Springs aquifer, western Oklahoma

The U.S. Geological Survey, in cooperation with the Bureau of Reclamation and the Oklahoma Water Resources Board, (1) quantified the groundwater resources of the Rush Springs aquifer in western Oklahoma by developing a numerical groundwater-flow model, (2) evaluated the effects of estimated equal-proportionate-share (EPS) pumping rates on aquifer storage and streamflow for time periods of 20, 40, and 50 years into the future, (3) assessed the uncertainty in the EPS scenario results, and (4) evaluated the effects of (a) projected groundwater-use rates extended 50 years into the future and (b) sustained hypothetical drought conditions over a 10-year period on stream base flow and groundwater in storage.

The Rush Springs aquifer is an important source of water for municipal and irrigation use by many communities and agricultural users in the study area. The study area is composed of about 4,970 square miles (3,181,003 acres) of Rush Springs aquifer bedrock deposits located in 14 counties. The study area also includes the alluvium and terrace deposits of the Canadian and Washita Rivers, as well as alluvium along the Little Washita River, Deer Creek, and a number of smaller tributaries of the Washita River that overlie the bedrock.

A numerical groundwater-flow model of the Rush Springs aquifer was constructed by using MODFLOW with the Newton solver. Groundwater flow was simulated for January 1979–December 2015 by using monthly stress periods, and an initial steady-state stress period was configured to represent mean annual inflows and outflows. The model was calibrated to groundwater-level observations at selected wells, monthly base flow at nine streamgages, stream seepage as estimated for the conceptual water budget, and Fort Cobb Reservoir stage.

The EPS scenarios for the Rush Springs aquifer were run for periods of 20, 40, and 50 years. The 20-, 40-, and 50-year EPS pumping rates under normal recharge conditions were 0.82, 0.49, and 0.43 acre-foot per acre per year, respectively. Given the 2,954,545-acre aquifer area used for the EPS scenarios, the 20-year rate corresponds to an annual yield of about 2,422,727 acre-feet per year. Groundwater storage at the end of the 20-year EPS scenario was about 13,321,000 acre-feet, or about 31,516,437 acre-feet (70 percent) less than the starting EPS scenario storage. This decrease in storage was equivalent to a mean groundwater-level decline of about 152 feet. Water availability under the EPS pumping rate was primarily from the western area of the model. Saturation was sustained though the entire EPS scenario where the aquifer was sufficiently thick or a shallow hydraulic gradient was present. Fort Cobb Reservoir stage was below the dead-pool stage after about 5 years of 20-year EPS pumping.

An uncertainty analysis was conducted to assess the uncertainty in the EPS scenario results. An ensemble of 400 random sets of possible parameter values was performed for the uncertainty analysis by using a multivariate normal distribution centered on the calibrated parameter values. The parameter bounds for the uncertainty analysis were determined by using the posterior covariance matrix, which allows for the incorporation of knowledge gained during the calibration process as well as observation uncertainty and the correlation between estimated parameters. The uncertainty results indicate a 95-percent confidence interval for the 20-year EPS pumping rate between 0.73 and 0.95 acre-foot per acre per year.

Projected 50-year pumping scenarios were used to simulate the effects of selected well withdrawal rates on groundwater storage of the Rush Springs aquifer. The effects of well withdrawals were evaluated by comparing changes in groundwater storage between four 50-year scenarios using (1) no groundwater use, (2) mean groundwater use for the study period (1979–2015), (3) increasing groundwater use, and (4) groundwater use at the 2015 rate. The increasing-use scenario assumed a 38-percent increase in pumping over 50 years on the basis of 2010–60 demand projections for western Oklahoma. Simulated groundwater storage changes ranged between an increase of 6.3 percent for the scenario with no groundwater use, and 0.9 percent for the scenario with 2015 groundwater-use rates. For the Fort Cobb Reservoir surface watershed, simulated groundwater storage changes ranged between an increase of 23.6 percent for the scenario with no groundwater use and a decrease of 4.0 percent for the increasing groundwater-use scenario. Groundwater-level changes were generally greater in areas with a large concentration of groundwater wells and groundwater use such as the Fort Cobb Reservoir surface watershed.

A hypothetical 10-year drought scenario was used to simulate the effects of a prolonged period of reduced recharge on the Rush Springs aquifer groundwater storage and Fort Cobb Reservoir stage and storage. Drought effects were quantified by comparing the results of the drought scenario to those of the calibrated numerical model. To simulate the hypothetical drought, recharge in the calibrated numerical model was reduced by 50 percent during the simulated drought period (1983–1992), and upstream inflows to the Canadian and Washita Rivers and associated tributaries were reduced by 37 percent. Groundwater storage at the end of the hypothetical drought period in December 1992 was about 42,983,000 acre-feet, or about 3,525,000 acre-feet (7.6 percent) less than the groundwater storage of the calibrated numerical model. This change in groundwater storage is equivalent to a mean groundwater-level decline of 15.8 feet. Simulated mean base-flow declines at the Canadian and Washita River streamgages were between 39 and 59 percent during the drought period. The minimum stage in Fort Cobb Reservoir at the end of the hypothetical drought period was 1,311 feet, indicating a storage capacity of only 10 percent of active conservation pool storage. The Fort Cobb Reservoir storage declines mostly resulted from reduced base flows in Cobb, Lake, and Willow Creeks upstream from the reservoir.