The Big Sioux aquifer in Brookings, Deuel, and Hamlin Counties, South Dakota, has been extensively developed and in some areas discharge, principally by wells, from the aquifer may be exceeding recharge to the aquifer.

A finite-difference method digital model was used to simulate steady-state conditions of the Big Sioux aquifer. Average annual water levels in the Big Sioux aquifer and base flow discharge of the Big Sioux River near Brookings for 1970 through 1976 were used in the model. The model-computed water levels were within a few feet of the actual annual average water levels and the computed annual average base flow was 66 cubic feet per second compared to the actual base flow of 58 cubic feet per second.

The computer model was used to model transient conditions by simulating water levels and base flow from April through August 1976 and comparing the results with actual data. Evapotranspiration and pumpage changes were made for each month. There was no recharge from precipitation during the test period.

Several different computer simulations were made using different estimates of hydrologic parameters and conditions. Specific yield was increased from 10 to 15 percent which resulted in a much greater base flow for each month. Effective depth of evapotranspiration was changed from 5 to 10 feet which resulted in a very large decrease in base flow. A computer simulation made without irrigation pumpage resulted in an increase in the base flow from 0.66 to 9 cubic feet per second for August 1976 in the Big Sioux River near Brookings. The actual base flow for August 1976 was .01 cubic foot per second.

A water budget showed 22.2 inches of precipitation (average annual), 0.65 inch of surface runoff, 1.06 inches of ground-water outflow (base flow to river, 1970-76), and 20.49 inches of evapotranspiration.

The water from the Big Sioux aquifer is a calcium bicarbonate type and specific conductance ranged from 407 to 1,790 micromhos per centimeter at 25°C. The water is generally very hard, having a mean of 454 milligrams per liter of hardness.

A model simulation using all the pumpage that would be allowed by irrigation permits approved as of February 1979 simulated the withdrawal of 43,900 acre-feet of water for about 4 months during which time there was no recharge from precipitation. If there had been no pumping for that period, evapotranspiration would have been 7,800 acre-feet more than occurred under pumping conditions and discharge to streams would have increased by 3,600 acre-feet.