Laserfiche WebLink
<br />3-03 <br /> <br />cloudburst flOod. . Where two or IOOre types of floods are distinct . <br />and do not occur prE"'lnn.1n""tlyinmutwiJ. caubinations,they should <br />not be combined. into asin8J,eseriesfor frequency C1&lysis.'It,c <br />is usus.lJ.ymore reliable in such cases to segregate the data in <br />accordance with type ilIld to combine o~ the final Cl1r'f'eS, :11'neces- <br />sary. In the Sierra Nevad.aregion of California end Nevada,fre- <br />quency .studies .are made se;pe.re.tely for re.ini'loods, 1Ih1ch occur prin- <br />cipBny dUr1ngthe ._ths of N~er through Karch, end for snowmelt <br />floods, which occur during thellXlD.ths' of A:pr1J. through Jul:r.FIOll's <br />for each of these two seasons are segregated strictly by cause'_... <br />those predominantly caused by snowmelt end those pre"nminA.ntly caused <br />byra:1n. In desert regiau, s11llliler thiD:lderstorma should be~lU!led <br />from frequency studies of rlnter rainfloods or spring snowmelt. floOds <br />and should be considered se:parate1y. <br /> <br />~. Occasionally a runoff' record ~ be interrupted by a period <br />of one or moreyeara. If' the interruption is caused by desttuction <br />of the gaging station by a large flood, tsUure to fill in the record <br />for that flood would have a biasing effect, Ybich should be aVoided. <br />However, if the cause of the interruption is known to be independent <br />of .flow magnitude, the entire period at interruption shOuIdbe elimi- <br />nated from the frequency arra;:r, since no bias would result. KnOll"l- <br />edge we m:l.ght h&Ve about floods observed at other locations 'during <br />periods of no record a.t the. site concerned can be utiJ.1zedas dis:.. <br />cussed in section 5. In .cases where no runoff records are &va1lable <br />on the .stresm concerned, it is usually best to estimate the frequency <br />curve asaw'hole using regional generalizations' discussed in .section <br />7, instead of attempting toest1.Jllate a canplete -series ofill.d1'\'tdwiJ. <br />. floods, because ordinarymethods-of estilMting iDdi Vidllal floods <br />te <br />tion) . ., <br /> <br /> <br />3-04. <br /> <br />ADJUSTMENT TO UNIFORM COflDrrICti . <br /> <br />Since frequency anal)'llis of hydrolOgiC data. is based on the <br />assum,ption of random occurrences, each item of data. must have occurred <br />under simils.r hydroJ.ogic conditions or mUst be lIdJustedto a. standard <br />uniform condition. If' controJ. by reservoirs or diversion for irri- <br />gation, etc., has s.ffected the runoff', some lIdJustment of the data. <br />is usuaJJ.y necessary. Where it is feasible to adjust to natural con- <br /><litions, it is adVisable to do so in order that the data will more <br />nearly conform to theoretical freqUency functions that have been <br />found to describe the frequency of natural hydrologic events. This <br />is accolll;pl1shed by standard routing procedures, and in JlIBDy cases <br /> <br />- 11 - <br />