FLOOD10717

• County: Mesa
• Community: Mesa County and Unincorporated Areas
• Title: FIS - Mesa County and Unincorporated Areas
• Date: 7/15/1992
• Prepared For: Mesa County and Unincorporated Areas
• Prepared By: FEMA
• Floodplain - Doc Type: Current FEMA Regulatory Floodplain Information

pond upstream. The city used the COE HEC-1 rainfall-runoff
computer program (Reference 13) to compute the lOa-year discharge.
The same model was used in this Flood Insurance Study with
appropriate precipitation values for the 10-, 50-, and SOD-year
storms.

The HEC-I computer program utilizes as input the following data:
rainfall, loss rates, unit hydrographs, base flow, and routing
criteria. Rainfall computations for the 10- and SO-year storms
were based on precipitation-frequency data and areal reduction
factors derived from the National Oceanic and Atmospheric
Administration Atlas. To obtain the precipitation value for the
SOD-year flood, a frequency-precipitation curve for the stream was
plotted using the 10-, 50-, and IOO-year value, and the curve was
extrapolated to a SOO-year value. Unit hydrographs were derived by
the Snyder Method and loss rates were established by hydrologists
familiar with the area. A base flow of 60 cfs was used.

Results of hydrologic analysis for Big Salt Wash and Little Salt
Wash presented in the 1976 COE Flood Hazard Information Reports
(Reference 15) were used in this study.

Plateau, Grove, Buzzard and Roan Creeks, among tributaries to the
Colorado Ri ver from the vicini ty of DeBeque to Glenwood Springs,
Colorado, are streams that flood from snowmelt as well as from
summer rain. Peak di scharges for 10-, 50-, 100-, and SOO-year
floods were determined using a regional floodflow frequency
analysis (Reference 16) based on rainfall and snowmelt data. The
results of the snowmelt and rainfall frequency analysis were
combined statistically to give a maximum flow for both events.
Regional curves of drainage area versus discharge were computed for
the four frequency floods for Roan Creek. Streamflow data of
rainfall flood events were insufficient to accomplish this,
therefore the SCS TR-20 computer program was used to simulate
rainfall flood peaks. The model was used on 16 watersheds of
varying sizes. A regional curve of drainage area versus peak
di scharge and frequency was developed for rainfall flooding. The
TR-20 analysis included the standard SCS Type II (24-hour) rainfall
distribution and curve numbers for an average antecedent soil
moisture condition (AMC-II). The discharge-versus-drainage-area
data from this analysis were plotted, and a regression line
computed for several frequencies.

A regional curve was developed for snowmelt flood events using
streamgage data from 8 streamgages in the area. The log-Pearson
Type III frequency distribution as defined in WRC Bulletin 17-8
(Reference 12) was used wi th a regional skew weighted wi th each
computed station skew. The data were plotted and frequency lines
were drawn.

The two regional discharge frequency-drainage area curves (rainfall
and snowmelt) were combined using a standard probability equation:

P(comb)=P(snow)+P(rain)-(P(snow)xP(rain))

21