Three subbasins in the West Branch Brandywine Creek watershed were modeled by the Dawby, Schaake, and Alley distributed routing rainfall- runoff model. The Honeybrook subbasin could not be calibrated because of nonrepresentative rainfall data. The Coatesville subbasin was calibrated, but not verified; the average standard error of estimate is 34 percent for peak discharge. The Modena subbasin was calibrated and verified. Average error for peak discharge is 38 percent for calibration and 24 percent for verification.
Predictive simulations using selected storms were made to determine the effects of projected population and four proposed flood-control structures in the Coatesville and Modena subbasins. Simulation of projected population growth in both subbasins showed that runoff volumes would generally have a greater percentage increase for low-magnitude floods than for high-magnitude floods. Proposed flood-control structure PA-436D had the greatest reduction in peak discharge at West Branch Brandywine Creek at Coatesville when a large quantity of runoff entered the Coatesville subbasin from upstream. WA-2 had the greatest reduction in peak discharge when most of the runoff came from within the Coatesville subbasin. Although WA-2 and PA-430 will control about the same drainage area, WA-2 causes a greater reduction in peak discharge at Coatesville because PA-430 is to be located upstream from Rock Run Reservoir, which provides some flood- water storage. In the Modena subbasin, the simulated effect of proposed flood-control structure PA-428 showed that it would lower the peak discharge of Sucker Run at State Route 82 and either raise or have no effect on peak discharges at West Branch Brandywine Creek at Modena. Most simulations of the effects of proposed flood-control structures in the Coatesville subbasin on peak discharge at Modena showed a reduction in peak discharge.
