FLOOD00403

• County: Statewide
• Title: Hydrology of a Highway Stream Crossing
• Date: 10/1/1979
• Prepared By: USDOT
• Floodplain - Doc Type: Educational/Technical/Reference Information

e

.

'.

e

,.

.
.

e

the fifteenth year (10 years after the occurrence of the 12,000 c.f.s.
peak) was only 700 c. f. s. The third 10 year period might have included
one max1IIIum annual peak of 20,000 c.f.s. and another of 18,000 c.f.s.
By this time we wouJ.d probab~ have arrived at the correct conclusion
that individual peak. rates of runoff are not spaced equally in t1me.

Average Frequencies and Average Recurrence Intervals

So far we have dealt with only a tiny segment of the life history
of the strearn anu thus have been able to examine only a small sample of
the aJmost illfini te number of possible combinations of maximum annual
peak rates. When we think of the stream as having been flowing for a
thousand years or more, it no longer seems surprisin<! that the only
co~clusion concerning frequency that we were able to draw from 30 years
of recora _s tLat peak rates of e:rry selected magnitude are not spaced
eqtlaD.y in ti:ne.

If we haa a 1000 year streamflow record available and were to
plot the maxl.'!lUlll annual peaks on graph pa.per, it ..ould probably look
ver: si:lilar to fieure 2. Of' course, no such 1000 year record exists
a.r.d figure 2 is merely a. synthetic presentation of one of the almost
u!'..li1:J. ted nwaber of possible coL1binations of rr.a.:d.mUln annual peak rates
tha~. could occur in a lOOO-year period. For each lO-,year period in thi 5
.;raph, the ',lBJCimum annual peak rates equal to or greater than 13,000
c.f.s. have been plotted 0.0; though they had occurr<;d ir:. ,hoe mi<1<i:..e 00:
the period. If.e were to cou.nt the number of peaks that ..ere equal
to cr grea+er than 13,000 c.r.s. we wouJ.d find that there were l~ such
peaitS in the first 100 ,'ear::; or an average of one in 10 :J<-ears. If,
however, 1o'e were to rr.a.ke our count for the period of year 50 to year
1~0 we wouJ.d find that -t..here was only six and that the number for the
period year 150 to year 250 was 14. Thus we see that average occurr-
ences obtained from 100 year records wouJ.d not give us consistent results
since they could be expected to vary from once in 7 years to once
in 17 years. If our averages had been obtained froc-. consecutive 200
year periods, however, the variation wouJ.d: have been from once in 8 years
to once in 11 years and when obtained from 5oo-year periods the averages
"'oulli have been a consistent once in 10 years.

Although it is true that no recurrence interval can be assigned
to a solitary peak rate of runoff, the above example demonstrates that,
for a long period of time an a.verage recurrence interval exists for
peaks that are equal to or greater than any selected minimum.

3-3