FLOOD10102 - Laserfiche WebLink

Home Browse Search

Flood Volume As previously discussed, the rainfall-runoff model produces both peak dis- charge and flood volume during the synthesis phase. A simple linear regression of the volume associated with each peak discharge was completed to determine if the synthetic data might yield a relation similar to that developed from the re- corded data (equation 1). This analysis was accomplished using data from first only the "Pueblo syntheses," then only the "Amarillo syntheses," and finally the combined syntheses ("Denver syntheses" were not used as previously discussed). The "Pueblo syntheses" yielded 1,044 floods ranging in peak discharge (Qp) from 3 to 12.900 ft3/S and flood volume (V) from 0.98 to 1,884 acre-ft. For lhe 521 floods from the "Amarillo syntheses" peak discharges ranged from 5 to 15,680 ft3/S and flood volumes ranged from 0.98 to 3,893 acre-ft. The following rela- tion s were determi ned: Pueblo data: V=0.141Qp 0.919 (5 =62, R=O.92) (4) e Amarillo data: V=0.139Qp 0.964 (5 =56, R=0.93) (5) e Combi ned data: V=0.123Qp 0.958 (5 =62, R=0.93) (6) e Of these three equations, the relation based only on the "Pueblo syntheses" (equation 4) is probably most representative of the Arkansas River basin in Colorado and results in estimated flood volumes about 23 percent higher than the relation based on the recorded data (equation 1). Compared with equation 1 results, the equation developed from the "Amarillo syntheses" (equation 5) yields volume estimates twice as great while the combined equation (equation 6) gives results 30 to 60 percent greater. REGIONAL FLOOD-FREQUENCY ANALYSIS Very seldom is flood-frequency information required at or near a gaging station where data are available for a station flood-frequency analysis. More typically, particularly for small drainage basins in the arid West, data of this type are required at an ungaged site. As a result, station flood-frequency information is many times regionalized (areally extrapolated) for planning and design purposes. For the Arkansas River basin in Colorado, examples of recent regional flood-frequency analyses include Patterson (1964) and McCain and Jarrett (1976). For this study the results of two regionalization methods are discussed in the following sections: multiple-regression analysis with basin characteristics and station-year analysis of recorded floods. i . Multiple-Regression Method One of the most effective ways presently known for defining streamflow characterisitcs on a regional basis is to relate them to basin and climatic charact- eristics by use of multiple-regression techniques applied to past data (Benson and Carter, 1973). Multiple-regression analysis of flood data generally includes log transformation of the data and, therefore, results in an equation of the form 17