This report describes the results of a study by the U.S. Geological Survey, in cooperation with the Delaware River Basin Commission, to develop a regional ground-water-flow model of the French Creek Basin in Chester County, Pa. The model was used to assist water-resource managers by illustrating the interconnection between ground-water and surface-water systems. The 70.7-square mile French Creek Basin is in the Piedmont Physiographic Province and is underlain by crystalline and sedimentary fractured-rock aquifers. Annual water budgets were calculated for 1969-2001 for the French Creek Basin upstream of streamflow-measurement station French Creek near Phoenixville (01472157). Average annual precipitation was 46.28 in. (inches), average annual streamflow was 20.29 in., average annual base flow determined by hydrograph separation was 12.42 in., and estimated average annual ET (evapotranspiration) was 26.10 in. Estimated average annual recharge was 14.32 in. and is equal to 31 percent of the average annual precipitation. Base flow made up an average of 61 percent of streamflow.

Ground-water flow in the French Creek Basin was simulated using the finite-difference MODFLOW-96 computer program. The model structure is based on a simplified two-dimensional conceptualization of the ground-water-flow system. The modeled area was extended outside the French Creek Basin to natural hydrologic boundaries; the modeled area includes 40 square miles of adjacent areas outside the basin. The hydraulic conductivity for each geologic unit was calculated from reported specific-capacity data determined from aquifer tests and was adjusted during model calibration. The model was calibrated for above-average conditions by simulating base-flow and water-level measurements made on May 1, 2001, using a recharge rate of 20 in/yr (inches per year). The model was calibrated for below-average conditions by simulating base-flow and water-level measurements made on September 11 and 17, 2001, using a recharge rate of 6.2 in/yr. Average conditions were simulated by adjusting the recharge rate until simulated streamflow at streamflow-measurement station 01472157 matched the long-term (1968-2001) average base flow of 54.1 cubic feet per second. The recharge rate used for average conditions was 15.7 in/yr.

The effect of drought in the French Creek Basin was simulated using a drought year recharge rate of 8 in/yr for 3 months. After 3 months of drought, the simulated streamflow of French Creek at streamflow-measurement station 01472157 decreased 34 percent. The simulations show that after 6 months of average recharge (15.7 in/yr) following drought, streamflow and water levels recovered almost to pre-drought conditions.

The effect of increased ground-water withdrawals on stream base flow in the South Branch French Creek Subbasin was simulated under average and drought conditions with pumping rates equal to 50, 75, and 100 percent of the Delaware River Basin Commission Ground Water Protected Area (GWPA) withdrawal limit (1,393 million gallons per year) with all pumped water removed from the basin. For average recharge conditions, the simulated streamflow of South Branch French Creek at the mouth decreased 18, 28, and 37 percent at a withdrawal rate equal to 50, 75, and 100 percent of the GWPA limit, respectively. After 3 months of drought recharge conditions, the simulated streamflow of South Branch French Creek at the mouth decreased 27, 40, and 52 percent at a withdrawal rate equal to 50, 75, and 100 percent of the GWPA limit, respectively.

The effect of well location on base flow, water levels, and the sources of water to the well was simulated by locating a hypothetical well pumping 200 gallons per minute in different places in the Beaver Run Subbasin with all pumped water removed from the basin. The smallest reduction in the base flow of Beaver Run was from a well on the drainage divide between the French Creek Basin and the Marsh Creek Basin to the south; the simulated base flow of Beaver Run at the mouth was reduced 1 percent. The greatest reduction in the base flow of Beaver Creek was from a well close to Beaver Run; the simulated base flow of Beaver Run at the mouth was reduced 8 percent. The simulations showed that (1) if the contributing area of a well is in a basin, pumping will affect stream base flow and water levels in that basin whether the well is inside or outside that basin; (2) wells in different areas of a basin away from a divide produce a similar reduction in base flow; (3) a well within a basin will derive more water from diverted base flow and less water from storage than a well on or near a basin divide; and (4) the reduction in base flow at the mouth of the stream is the same for a well in the headwaters and a well downstream near the confluence.

Model simulations illustrate some of the typical analyses and results that can be produced. The model was calibrated using annual values for recharge and ground-water ET and then was run using the annual values in a seasonally independent transient mode to show changes with time. The timing and relative magnitude of some of the changes simulated with the model when viewed in terms of a normal climatic year may be subject to considerable uncertainty because of the variability in seasonal recharge and ground-water ET rates. Transient model simulations for short-term periods are indicative of possible hydrologic system response and are considered an approximation.