A flood-estimating method is presented which applies to drainage basins in Delaware without urban development and covers selected recurrence intervals from 2 to 100 years. The method was developed by multiple-regression techniques. The State is divided into two regions and sets of equations for calculating peak discharges based on physical basin characteristics are provided for each region. The boundary between regions generally corresponds with the division between the Piedmont and Coastal Plain provinces. In the northern region, flood-peak discharges were related to basin drainage area and storage. In the southern region, flood peaks were related to drainage area, slope, storage, forest cover, and two composite soil categories. Standard errors of estimate for the regression equations in the northern region ranged from 30 to 39 percent. For the southern region, the standard errors of estimate varied from 38 to 40 percent. Without using the two soil parameters in the southern region, the standard errors of estimate varied from 57 to 70 percent. Annual flood peaks, basin characteristics, and flood-frequency distributions are tabulated for the 60-gaged sites used in the regression analysis. At 23 of these sites, a rainfall-runoff model generated additional flood-peak data which were used in defining the flood-frequency distributions. (Woodard-USGS)
Additional publication details
USGS Numbered Series
Technique for estimating magnitude and frequency of floods in Delaware
Water-Resources Investigations Report
U.S. Geological Survey,
iv, 69 p. :ill., fold. maps (3 fold. col. in pocket) ;27 cm.