|
<br />I ..
<br />;V<
<br />
<br />0008J5
<br />
<br />Mvl/.Ji\,'
<br />
<br />JUNE
<br />
<br />INTERNA llONAL SYMPOSIUM ON HYDROMETEOROLOGY
<br />",' AMERICAN WATER RESOURCES ASSOCIATION
<br />
<br />'. :!:;L1982
<br />
<br />:,"1"
<br />
<br />. ~ . ..'.~. '..
<br />
<br />/
<br />
<br />'.;':' "': "~-:'.)) ~;:,.,:. ./-11': .!....:-.......
<br />-'
<br />
<br />THE loo-YEAR 24-HOUR EVENT: FACf OR FICTION
<br />
<br />Jerald S. Fifield'
<br />
<br />ABSTRACT: For the hydrologic engineer, it is important that peak
<br />flows resulting from a 24.hour climatic event 1;>e representative of
<br />the area under investigation. An important paramelcr for deter.
<br />m.ining pe3k now (and in generating the accompanying hydrograph)
<br />from an ungaged W3tershcd is utilizing:ln ;lccuratc time distribution
<br />for the design storm. Frederick, ~l al.(l981), developed a method
<br />for determining how precipitation is distributed wilh respect to
<br />time for 'various areas throughout the western United Stales. By
<br />comparing lOQ.yc.:u 24.hour flood hydrographs simulating no~
<br />from ungJged watersheds .using the Frederick, et at. (1981), distribu-
<br />tion to the Soil Conservation Service Type I or Type II distribution,
<br />it become~ e"ildent that serious hydrologic CHon could de'Ve\op for
<br />which the hydrologic engineer would be responsible.
<br />(KEY TERMS: precipitation; nood hydroglJphs; rainfall distribu-
<br />.tion.)
<br />
<br />PROBLEM
<br />
<br />For the hydrologic engineer, regulations that govern the
<br />design of sedimentation ponds, development of diversion
<br />structures, culvert sizing, and so f~rth, often require an
<br />accurate determination of flood flows from ungaged w3ter-
<br />sheds resulting from a 100.year 24-hour climalic evenL
<br />k analyses are undertaken, it becomes evident that the
<br />concept of a 24-hour event is not fully understood by the
<br />engineer nor the time distribution of precipitation accurate-
<br />ly represented. Consequently, results may be in error caus-
<br />ing excess expenditures or creating situations where sub-
<br />STantial damage may occur if the lOO-year 24.hour event
<br />becomes a reality.
<br />
<br />THE 24-HOUR EVENT
<br />
<br />In parts of the nation, a single storm may have a duration
<br />lasting 24 hours. However, for the weslern United St:Jtes a
<br />storm lasting 24 huurs is not a common occurrence (except
<br />for Ihe Pacific Northwest) and usu,lIy not of the type thaI
<br />produces fiood Ilows. Retcher, er aI. (1981), observed in
<br />Iheir studies of precipil:Jtion events for ,enfral Utah only
<br />4 percent uf th~ :ilurrns recurded had durations longer than
<br />
<br />I Presld..:nl, HydroDynamks Inc.. P.O. Box 1327, P:ukcr, Colorado 80134.
<br />
<br />7 hours and rarely produced floods in the study area. How-
<br />, ever. they found [hat about 90 percent of the stormshad a
<br />dura tion of 4 hours or less.
<br />The U.S. Weather Bureau (1963) indicates that the
<br />months having maximum recorded 24-hour tainfall for the
<br />weslern United States are Morch through November except
<br />for areas west of the Sierra Nevada and Cascade Mountains.
<br />Review of climatic records further illustrate that maximum
<br />recorded point rainfall for the hours of 1, 2, 3, 6,12, and
<br />24 have a greater Occunence during the months of May
<br />through August when thunderstorm aclivity is more likely
<br />to occur. Since thunderstorm activities are usually of short
<br />duration, a 24-hour event may consist of Jl10R than one
<br />dimalic event. Hence, a suggested definition for hydrologic
<br />engineers to use concerning a 24.hour climatic event during
<br />the thunderstorm season may be "the amount of precipita-
<br />tion that has accumulated al a point for 24 hours following
<br />the first measurable rainfall."
<br />
<br />RAINFALL DISTRIBUTION
<br />
<br />Perhaps more important to the hydrologist predicting
<br />flood flows from ungaged watersheds is how the distribulion
<br />of rainfall will occur over the area being studied. A set of
<br />widely used dislributions were developed by the Soil Con-
<br />servation Service (SCS) using the U.S, Weather Bureau rain-
<br />fall frequency atlas, namely Type I ,nd Type II diHricu-
<br />tion. The Type I dislribution is designed for Hawaii, Alaska,
<br />and the Coastal side of the Sierra Nevada and Cascade
<br />Mounl.ins in California, Oregon, and Washington. ,The
<br />Type II distribution was intended to be used in the re-
<br />mainder of the United States, Puerto Rico, ,nd the Virgin
<br />Islands. Figure I illustrates these distributions.
<br />, Since thunderstorm activities can result in high intensity
<br />precipitation for short durations and may create destructive
<br />flood flows, one may question whether a Type II distribu-
<br />tion is adequate for most of the western United States, To
<br />further analyze this problem, Frederick, er aI. (I 981),
<br />
<br />531
<br />
<br />.~'~~:~t~~~
<br />.. :::~~~~.
<br />
<br />....~., ...-
<br />......,..,.
<br />
<br />'. H
<br />. .'.....:.
<br />
<br />:.,:
<br />
<br />f'
<br />~
<br />f!,
<br />I
<br />I
<br />
<br />"
<br />i
<br />
<br />"
<br />I
<br />
<br />fl
<br />I
<br />
<br />,
<br />,
<br />
<br />,
<br />
<br />,
<br />
<br />r,
<br />~'
<br />
|