Actual Storm Events Outperform Synthetic Design Storms: A Review of SCS Curve Number Applicability

Two historic storm events occurred July 8, 1999, and August 19, 2003, over Las Vegas, Nevada, that met or exceeded 100-year frequency values with regard to rainfall depths in some areas and peak runoff levels in others. Although devastating in terms of loss of life and property damage, the two storm events provide a rare opportunity to compare synthetic design storms to actual rainfall events in the arid Southwest. Each storm, which occurred over a distinct watershed, is reconstructed using real-time data from a rain gage system maintained by Clark County Regional Flood Control District. Actual rainfall intensities are then compared to a standard 100-year, 6-hour rainfall distribution. The difference in "real" and synthetic storm intensities offers insight into why established 100-year rainfall depths and flowrates were exceeded during these two events. Actual precipitation depths and intensities are then input into HEC-1 Flood Hydrograph Package models to evaluate the sensitivity and applicability of SCS (Soil Conservation Service) curve number values for estimating runoff from developed and undeveloped watersheds in an arid environment. By comparing runoff levels recorded at stream gages in each watershed, this paper will show that SCS curve numbers effectively represent runoff from developed watersheds and overestimate undeveloped watershed runoff in an arid environment.