FLOOD05157 - Laserfiche WebLink

Home Browse Search

..~~ ~_.", .=lIIiIiIti.lII-J"""ilt-~1 -.miiIiilIiii""......,&'~,',",;,'; MANUAL FOR ESTIMATING FLOOD CHARACTERISTICS ~ Annual peak discharges through September 30, 1973, for each 9a~ing sta- tion used in this study were fitted to the log-Pearson Type I I I distribution. A subjective appraisal of high and low outl iers was made based on tre rea- sonableness of the computed flood-frequency curves. High outliers cause frequency curves to estimate extremely large flood discharges, especially for the larger recurrence intervals, through an increased standard deviation and large positive skewness. Low out! iers cause large negative skew"ess but in- crease the standard deviation of the frequency distribution. To 5or,e degree, these characteristics are compensating, but extreme negative skewness can produce flood-frequency relations which estimate little or no increase in discharges with increases in recurrence intervals. The outl iers ......'ere el im- inated, where warranted, and the flood characteristics were recomputed based on the remaining annual peak discharges. The gaging stations where outl iers were el iminated are footnoted in tables S-B. The computed peak discharges for the 10-, 50-, 100-, and SOD-year recur- rence-interval floods for each gaging station used in this analysis are listed in tables 5-B. COMPUTATION OF FLOOD DEPTHS Depths for the 10-, 50-, 100-, and SOO-yea r floods we re computed for 223 gaging stations in Colorado and adjacent States. Some of the stations used in the discharge analysis were not used in the depth analysis because of difficulties in obtaining stage-discharge relations or because of suspected inaccuracies in extending some stage-discharge relations to cover extremely large. flood discharges. :~ The most recent stage-discharge relation for each station (usually through September 30, 1973) was used as the base relation, with previous relations being used where applicable to extend the base relation both up- ward and downward. Many of the relations required long extensions in order to cover the range in discharge. Straight-line extensions on log-log graph paper were made except where other information, such as bankfull stage, indicated curvature. The depth computations consisted of picking a stage for the 10-, 50-, 100-, and 500-year discharges from the stage-discharge relation. Next, the stage of the point of zero flow was obtained from the relation and subtracted from the stage of the flood discharge. Selection of the point of zero flow is illustrated in figure 12. The computed depths for the 100-year flood are listed in tables 5-8.