FLOOD00918

• County: Larimer
• Community: Fort Collins
• Title: Precipitaion Frequency Study
• Date: 7/16/1998
• Prepared For: Fort Collins
• Prepared By: WRC Engineering, Inc.
• Floodplain - Doc Type: Floodplain Report/Masterplan

I
I
I
I
I
I
I
I
I
II
I
I
I
I
I
I
I
I
I

2. Two Hour Analyses:
Currently, the City of Fort Collins utilizes a two hour design storm for the design of drainage
systems in small watersheds. Therefore, a comparison of the resulting 2 hour event data is
needed. A direct comparison to NOAA Atlas was not possible in this case since an equivalent
period of record, as was used with NOAA Atlas 2. for the period prior to 1948 was not
available. Therefore, the analyses were based on the periods ofrecord from 1948 to 1996 and
1997. The results are presented in Table 3 and Figures 6 and 7. The supporting analyses are
contained in Appendix O.

As with the 24 hour analyses, the Fisher-Tippet distribution tends to replicate the NOAA Atlas
2 Fisher-Tippet distribution better, but does not fit the data with the 1997 storm as well as the
Log Pearson Type III distribution (Appendix 0). The difference with the 2 hour analysis is
that it suggests that NOAA Atlas is overestimating the 2-, 5-, and IO-year return periods, and
underestimating the 25-year return period and above for the 2 hour analysis.

B.

CSU ANALYSES:

The analyses contributed by Oli Sveinsson at Colorado State University were done using the same data
periods as those used by WRC. Oli's analyses however, focused on the hourly data. "at-site" and
regional, using GEV and LN-3 data distributions weighted by L-moments rather than conventional
moments. Both of which give a reasonable fit to the data points used. Oli's results are shown in table
4 and have been converted from annual series to partial duration where needed. For more information
on these analyses, consult appendix F. Oli recommends using a regional GEV analysis of the 2 hour
data that includes the 1997 storm as recorded.

Comparing this analysis with WRC's 2 hour Log Pearson III analysis. shows some differences. Figure
8 shows that in general. the WRC results match for the 2-year return period, but are significantly
higher for the rest of the return periods, with the lOO-year return period being 0.69 inches higher.
When WRC's results are compared with Oli's "at-site" analysis, the values are quite similar throughout.
Therefore it can be concluded that the differences are attributed mostly with the "at-site" versus
regional analysis difference. When discussing the regional analyses. Oli states in his report, "This
should indicate that Fort Collins is an extreme compared to the other two stations used in the regional
analysis", recommending to stilI use the regional analysis, most likely due to having only 50 years

-7-