The ground-water resources of the White River basin in and near Madison County, Indiana, were investigated by mapping the aquifers, estimating their hydraulic properties, determining the distribution of potentiometric head in the aquifers, and estimating some of the components of the ground-water budget from data collected in the field. This information was used to construct and calibrate a five-layer, digital ground-water flow model. The model, constructed and calibrated to ground-water-level and seepage data collected during the study, simulates conditions for the autumn of 1976. The model was used to provide estimates of ground-water potential in terms of yield, drawdown, and depletion in streamflow.

Glacial drift covers nearly the entire study area and ranges in thickness from 0 to approximately 300 feet. Beneath the drift lie Ordovician to Devonian limestone, dolomite, and shale. A bedrock valley trends west-northwest through the center of the study area. Relief of the bedrock surface is approximately 325 feet.

Four confined sand and gravel aquifers interbedded in the glacial drift, a bedrock aquifer, and an unconfined outwash aquifer are the three most important aquifer systems in the study area. The nearly horizontal, areally discontinuous, confined sand and gravel aquifers have an average thickness of about 15 feet and an average hydraulic conductivity estimated to be 433 feet per day. The bedrock aquifer underlying the entire study area has a thick-ness estimated to be 150 feet and an average transmissivity estimated in previous studies in the basin to be 1,340 square feet per day. The unconfined outwash aquifer, although thin and narrow, is an important aquifer because of its proximity to sources of induced recharge.

Water-level fluctuations in observation wells in the basin indicate that the ground-water system is in dynamic equilibrium. Ground-water seepage to streams at 90-percent flow duration on October 1, 1976, was estimated to be between 15.9 and 74.4 cubic feet per second. Ground-water pumpage for 1976 was estimated to be 19.7 million gallons per day (30.5 cubic feet per second). The water budget, as simulated in the model, indicates that the rate of inflow to the ground-water system in the modeled area is 92.4 cubic feet per second. Of this, 78 percent is from effective areal recharge of precipitation, and 22 percent is from ground-water flow across the boundaries into the study area. Thirty-three percent of the ground-water outflow is pumpage, 59 percent is seepage to streams, and the remaining 8 percent is ground-water flow across the boundaries out of the study area.

Model simulations of eight pumping plans provide a general assessment of the water-yielding potential of the three major aquifer systems and indicate that as much as 2.5 million gallons per day can be obtained from well fields about half a square mile in area. Model results also indicate that, for as much as 2.5-million gallons per day pumping, flow in the large streams will not be significantly affected, whereas flow in the small streams may be significantly affected. In addition, simulations indicate that adding more wells to the well field northwest of Anderson, Ind., would probably be no more advantageous than increasing the pumpage in the existing wells. How-ever, well interference, well hydraulics, and pumping-level constraints were not considered in the investigation of alternative methods of expanding the well field. In developing the ground-water system in the area, use of many small, scattered well fields that produce less than about 3 million gallons per day may be more favorable hydrologically than a few, heavily pumped well fields.