Groundwater-flow model and effects of projected groundwater use in the Ozark Plateaus Aquifer System in the vicinity of Greene County, Missouri — 1907-2030

Abstract

Recent and historical periods of rapid growth have increased the stress on the groundwater resources in the Ozark aquifer in the Greene County, Missouri area. Historical pumpage from the Ozark aquifer has caused a cone of depression beneath Springfield, Missouri. In an effort to ease its dependence on groundwater for supply, the city of Springfield built a pipeline in 1996 to bring water from Stockton Lake to the city. Rapid population growth in the area coupled with the expanding cone of depression raised concern about the sustainability of groundwater as a resource for future use. A groundwater-flow model was developed by the U.S. Geological Survey in cooperation with Greene County, Missouri, the U. S. Army Corps of Engineers, and the Missouri Department of Natural Resources to assess the effect that increased groundwater demand is having on the long-term availability of groundwater in and around Greene County, Missouri. Three hydrogeologic units were represented in the groundwater-flow model: the Springfield Plateau aquifer, the Ozark confining unit, and the Ozark aquifer. The Springfield Plateau aquifer is less than 350 feet thick in the model area and generally is a low yield aquifer suitable only for domestic use. The Ozark aquifer is composed of a more than 900-foot thick sequence of dolomite and sandstone in the model area and is the primary aquifer throughout most of southern Missouri. Wells open to the entire thickness of the Ozark aquifer typically yield 1,000 gallons per minute or more. Between the two aquifers is the Ozark confining unit composed of as much as 98 feet of shale and limestone. Karst features such as sinkholes, springs, caves, and losing streams are present in both aquifers, but the majority of these features occur in the Springfield Plateau aquifer. The solution-enlarged fracture and bedding plane conduits in the karst system, particularly in the Springfield Plateau aquifer, are capable of moving large quantities of groundwater through the aquifer in relatively short periods of time. Pumpage rates in the model area increased from 1,093,268 cubic feet per day in 1962 to 2,693,423 cubic feet per day in 1987 to 4,330,177 cubic feet per day in 2006. Annual precipitation ranged from 25.21 inches in 1953 to 62.45 inches in 1927 from 1915 to 2006 in the model area. Recharge to the model was calculated as 2.53 percent of the annual precipitation and was varied annually. Recharge was distributed over the model area based on land slope and was adjusted in the city limits of Springfield to account for the impervious surface. A groundwater model with annual stress periods from 1907 to 2030 was developed using a transient calibration period from 1987 to 2006 and a prediction period from 2007 to 2030 to simulate flow in the Springfield Plateau aquifer and the Ozark aquifer. For the model area of approximately 2,870 square miles, the model hydrogeologic units and hydraulic properties were discretized into 253 rows, 316 columns, and 3 layers with the layer boundaries crossing hydrogeologic unit boundaries in some areas. The horizontal cell spacing was 1,000 feet by 1,000 feet. The model was calibrated by minimizing the difference between simulated head and observed water levels and simulated and observed flows in rivers and springs. Population and the associated groundwater use were estimated for 12 communities and the unincorporated area of Greene County based on past growth. Each was analyzed individually, and a low and high annual rate of growth relative to the 2006 population was computed for each community or group. Low growth rates ranged from 0.215 percent per year in Springfield to 6.997 percent per year in Rogersville. Total growth from 2006 to 2030 at the low growth rate ranged from 5.2 percent in Springfield to 167.9 percent in Rogersville. High growth rates ranged from 0.236 percent per year in Springfield to 7.345 percent per year in Rogersville. Total growth from 2006 to 2030 at the high growth rate ranged from 5.7 percent in Springfield to 176.3 percent in Rogersville. Response of the flow system to selected hypothetical pumping stresses and recharge conditions was simulated using the calibrated model. Seven hypothetical scenarios were simulated from 2007 to 2030 to test the effects of various stresses on the head in the Ozark aquifer. Hypothetical scenario 1 continued the 2006 pumping rates without change to the end of 2030. Scenario 2 assumed a low population growth rate with a 4-year drought at the beginning of the prediction period. Scenario 3 assumed a low population growth rate with a 4-year drought at the end of the prediction period. Scenario 4 assumed a high population growth rate with a 4-year drought at the beginning of the prediction period. Scenario 5 assumed a high population growth rate with a 4-year drought at the end of the prediction period. Scenario 6 and 7 had one new industrial well installed within the city limits of Springfield and one new industrial well installed about 3.5 miles east of Rogersville. Scenario 6 assumed a low population growth rate and scenario 7 assumed a high population growth rate. Results were compared by examining differences in head at the end of the simulation period. All scenarios examined resulted in potentiometric-surface declines from 2006 levels. Results from scenario 1 indicated that even with no increase in pumping, the potentiometric surface in the Springfield area continued to decline. The maximum decline of approximately 62 feet from the 2006 potentiometric surface occurred in Springfield. The maximum decline from the 2006 potentiometric surface in scenarios 2 and 3 was approximately 203 feet and in scenarios 4 and 5 was approximately 207 feet. The drought occurring at the end of the simulation period tended to broaden the drawdown area relative to the drought at the beginning. Drought timing did not substantially affect the potentiometric surface in the Ozark aquifer except for where the Ozark aquifer was exposed. Although not a substantial difference, the high population growth rate scenarios tended to have larger declines than the low population growth rate scenarios. As in the previous scenarios, little difference was noted between the low and high growth rate in scenario 6 and 7. Scenarios 6 and 7 showed declines of more than 640 feet from the 2006 potentiometric surface at the new well located in Springfield. The drawdown at the new wells decreased relatively quickly with increased distance from the well. Simulated head in the nearby cities of Nixa, Ozark, and Republic was nearly the same for scenarios 2 through 7 and was lower than the head predicted for scenario 1. Results from scenarios 2 through 7 indicate that the potentiometric surface in 2030 near these cities could decline 100 feet or more from the 2006 levels. Because model layers 2 and 3, representing the Ozark confining unit and most of the thickness of the Ozark aquifer, were simulated as confined, drawdown in the wells in the area of the Ozark aquifer that is unconfined or becomes unconfined during the simulation period will likely be under predicted.