FLOOD10102

• County: Statewide
• Basin: Statewide
• Title: Rainfall-Runoff Modelling and Preliminary Regional Flood Characteristics of Small Rural Watersheds in Arkansas River Basin - Colorado
• Date: 1/1/1981
• Prepared For: CDOT
• Prepared By: USGS
• Floodplain - Doc Type: Floodplain Report/Masterplan

Although nonuniform rainfall distribution, rainfall intensity and duration,
and soil-moisture conditions result in variations in runoff volume for floods of the
same magnitude, a relation between peak discharge and flood volume was deter-
mined from the 61 largest flood-hydrographs recorded at the 17 sites. These
hydrographs were characteristically the result of short-duration, high-intensity
rainfall typical of convective storms or thunderstorms, but not typical of
frontal-type storms. For these events, peak discharges ranged from 53 to 3,300
ft3/S and flood volumes ranged from 3 to 150 acre-ft. The relation is:

V=0.112Qp 0.922

(S =60, R=0.79)
e

(1)

where:

V=flood volume, in acre-feet;

Qp=peak discharge, in cubic feet per second;
S =average standard error of estimate, in percent; and
e
R=correlation coefficient.

In a previous study of 105 floods on small Wyoming basins, Craig and Rankl (1978)
found the following similar relation:

V=0.131Qp 0.878

(S =55, R=0.90)
e

(2)

Analysis of Synthetic Floods

Accuracy in estimating the magnitude and frequency of floods at a site im-
proves with the number of years of flood information on which the analysis is
based. The time-sampling error associated with an estimate of the 100-year peak
discharge based on 10 years of station data is considerably greater than if it
were based on 50 years of station data. For example, Livingston (1970) found
that for streams in the mountainous region of Colorado the standard error for the
25-year peak discharge decreased from 24 percent with 10 years of record to 11
percent with 50 years of record.

To improve the flood-frequency relations for stations in this study, a rain-
fall-runoff simulation model, described in detail by Dawdy, Lichty, and Bergmann
(1972), was calibrated and used to synthesize a long-term annual flood series for
15 of the sites.

Description of Modeled Rainfall-Runoff Data

From all rainfall-runoff data collected at the 17 study sites, a data set
consisting of 141 floods was selected for use in the calibration of the rainfall-
runoff model. Selection of floods for inclusion in this data set was based on (1)
the relative compatibility of rainfall totals or intensities with runoff volumes or
peak discharges, (2) the reliability of recorded stage and rainfall data, (3) the

9