Laserfiche WebLink
<br />3-01 <br />Sm'l'IOO3. - FLOOD PEAK FRElf./mcX - GRAPHICAL METHOD <br /> <br />3-01. . <br /> <br />USE. Al'lD RELA'l'IVE ADVAN'l!AGES <br /> <br />Every f'reqUllllC7 study should be plotted gra:pb.ically, eTen <br />thouab the results can be obtained entirely ~ically as <br />descri~:in section 4, in order that obserTSd data~be <br />'Y1sual4r c~ with the deriTed cuneo The gr"aphic&1lllllthod. <br />of' f'requenq-e determination can be used f'or ~. type of' <br />frequenq st~, but ana.lyi;ical I118thode have certain ad:nntages <br />where they Sore applicable (see paragraph 4-(1). The principal <br />advantages of gnl.phicalmethods are that they are genere.1.4" <br />s;pplicable, that the derbed curve can be easily nsuaJ..1zed, 8ZId <br />that the observed data can be read14r com;pared with the cOl\Pl1ted <br />results. However, graJidcal methode of' frequency e.na.ly'sis are <br />inferior in accuracy to analytical methode where the. latter a:p>>1y, <br />8ZId do not provide means of' evaluating the reliability of the <br />est1lllates. Coinparison of the adopted curve with plOtted ~ints <br />is not an index of reliability as in correlation analysis, but it <br />is often erroneous4r assUllIed to be, thus 1m;ply1ng a IDI1ch greater <br />reliability than is actually attained. For these reuons, graphical <br />methode should be 11Di:l. ted to those cases where a.nalytical methodS <br />do not aPP4r (that is, where frequeriey cums are too irregular . <br />to compute analytically) IIIId to use as a nsual aid or check on <br />analytical canpo.tations. . <br /> <br />3-02.. <br /> <br />GENERAL PROCEDURE <br /> <br />!,. Graphical construction of a frequency Curfe s1mp4r con;" <br />dsts of' arranging the selected data in the order at magnitude IIIId <br />p].otting the magnitude of' each item on the vertical scale S€;&1nst <br />s es e exce. e requency 0 ng position - see p&ra- <br />graph 3-05) On the horizontal scale, using a suitable grid (see <br />paragraph 3-06). A smoOth Curfe drawn through the points is the <br />desired frequency curve. . <br /> <br />b . D&ta used in the coJistruction of frequency curves of <br />peak nOlI's cODllist of the -~""1J>I flOll' for each year of record <br />IIIId all at the secOlld&ry flQWs that exceed a selected base -.alue. <br />This. base -.alue must be. _'1"1' than any floodtlow that is at <br />iItqlortance iD, the aal:18is, lIZId should also be 1011' enough so that <br />the totaJ. nUlllberof ~lood.s in excess of' the baSe equals or exceeds <br />the nUlllber of years of record. Ordinarily, the latter criterion <br />controls, 8ZId the two series of _ts tabulated are equal in <br />nUlliber. Exhibit 3 is a s8ll\l1lAl talnllation of' data..d1rect4" frCIIII <br /> <br />- 9 - <br />