FLOOD10474 - Laserfiche WebLink

Home Browse Search

For the regional regression analysis, 15 gaging stations located in the drainage basins of North Fork South Platte River, Bear Creek, Clear Creek, Upper St. Vrain Creek, and Upper Big Thompson River were selected. These stations were selected so that peak flows are not significantly affected by diversions and/or regulations and lengths of record are at least 15 years. Among the 15 stations, the drainage areas range from 14.4 to 575 square miles and the mean watershed elevations range from 8,800 to 11,140 feet. The average length of record is 43 years. From the recorded streamflow charts, the annual peak flows of rain floods and snowmelt floods were determined for each gaging station. For each individual station, the mean, standard deviation (and hence the coefficient of variation), and skew were computed from the annual flood series for each type of flood event. Regressions of the means, coefficients of variation, and skews on drainage areas and mean watershed elevations were then carried out for the 15 stations in the region. CUHP was used to estimate the 10-, 50-, and 1 DO-year flood discharges for the study reaches of Mount Veroon Creek and Bear Creek Tributary No.7. The 500-year flood discharges were estimated by linear extrapolation of the flood-frequency curves based on the 10-, 50-, and 1 DO-year discharges. A detailed description of the CUHP is given in Volume I of the "Urban Storm Drainage Criteria Manual" (Reference 42). Peak discharges at selected locations in the study reaches were obtained by routing the flood hydrographs for each subbasin computed by the CUHP. The USACE, Omaha District, completed a Flood Plain Information report on Bear Creek and Mount Vernon Creek for the Town of Morrison in 1971 (Reference 51). The IOO-year discharge of 27,000 cfs was estimated for Bear Creek at Morrison using a procedure called "expected probability". Because discharges determined using this procedure are not accepted by FEMA, no consideration was given to using this value in the study. Study of the Morrison gage data resulted in a discharge of 13,500 cfs and 90 percent confidence interval between 7,730 and 28,300 efs for the 1 DO-year flood. Because no stream gage data were available for the streams studied within the city of Westminster, a rainfall-runoff analysis was conducted on the watersheds to determine the flood discharges. This was accomplished by using the CUHP rainfall-runoff computer program to develop the storm hydrographs (Reference 42) and the USACE HEC-l flood hydrograph package computer program for the stream and reservoir routings (Reference 52). For this analysis, basin characteristics that define the size, shape, and runoff characteristics of the watershed, as well as rainfall amounts based on the selected recurrence intervals (obtained from the NOAA Atlas of Precipitation (Reference 46)) were used to compute flood hydrographs for various design points in the basin. All stream and reservoir routings were accomplished using the Modified PuIs Method. Since there was a lack of 500-year precipitation data, the 50D-year frequency storm runoff values at each design point were calculated. The logarithmic values of the 10-, 50-, and IOO-year peak discharges were fit to a regression line by method of least squares. The 50D-year discharges were analytically extrapolated from the regression line based upon a log-normal probability relationship. For more detailed information see the final hydrologic analysis report for Big Dry Creek and Tributaries prepared by Greiner Engineering Sciences, Inc., January 1984. 29