|
<br />would provIde a means of d1~$e~tndttng warn1"g in add1tion to servtng
<br />for flood det"ctton. Only "lt~r"<lt1ve 6 "'ould not provide thh capa-
<br />b11ity. Alternative ~ would have only on~ d~tectinn n~twork, while
<br />the oth~r" ~ould pro~ld~ t"'o ~ltern~ttve means of flood detection.
<br />
<br />RELIABILITY A.~D CREDIBILITY
<br />
<br />altho<.l&h thc annual sprtng rhc frolll s"o",,"clt attracts .."Ole public
<br />attcnt ton. Staff gage" '.rould he located a~ or a ahort drtve from
<br />()b~arv~r residences. Although bi"oculars and <I ~potUght would be
<br />provlded, heavy ratn ~ay reduce visibtlity such that th~ gng~ canoo~ be
<br />read. Although qualitattve da~a abo<.lt river levels ~ould bc report~d,
<br />the prol:lptnes. of reports Tllight vary fro", one observer to another. The
<br />location of potentilll obsarvers WaS determined by fieid observation or
<br />by points ",here residen~~s ...ere indt~ated on USGS quadrangle llIaps. If
<br />It develOps that rhere are nO longer residents at .Ome of these 10~a-
<br />ttons, the ob~erver net"oI'k would not be dens"- "-""Ough.
<br />
<br />If cost and lead time are not .ufficient crIteria, select ton of a"
<br />aU..rnaUve '""y have to be toade on the ba.is of other perfor",ance
<br />criterta. The reporr "",de for the Boulder naod warni"S plan(25)
<br />evalu<ltedaUernativesua1ngcriterias"ch<lsrel1ah1lity, credib1lity,
<br />non-flaad "'<lrning beMf1ts, tmplemen.tatlon, ~nd flexibility. Reli-
<br />abIlity and credibiltty ",ere used in thb report. Reliabtl1ty 19
<br />deftned as the dependability at aperation of II ",arning syste~'s eo..-
<br />ponent parts. Credtb1lity ts d"fined a. the certainty of prediction of
<br />a flood. Lo" credlhtlity "'arnings produ~e excessive nll",bers of faiae
<br />alsrms or act<.laily ",Iss flood events.
<br />
<br />There sre $OIlle advantage~ to v19ulll observer reporta. These
<br />reports are probably ~orc believable than lIuto~atie r"-porrs, thereby
<br />in~pirtng more confidence to take actian. Volunteer ~bservers can give
<br />qualitatIv,,- data in addirion to ratnfall a~unt~, aueh as wind $peed,
<br />llghtaing tntenairy, snd changtng storm co~dfttona,
<br />
<br />It is difftClllt ta ca"'p<lre reltabiltty aud credibility of flash
<br />flood detection alternatives ~ith ~ixed methods of obaervation. R<lther
<br />thsn to auign subjecttve ratings to each "tter"attve, ouch a3 "fair"
<br />or "good", the ad\1<1ntagcsnnddl""dvnnt.,ges of co"-poncnts of the
<br />alternattves are outlined,
<br />
<br />Those alternlltives involving volunteer rain and stream gages are
<br />dependent on the readiness of observers, At the tillle of ~ nor,", sa,"e
<br />may he a>lay or asle~p. ~atnten,'nce of the rain gages and c~d!os is
<br />"Iso II. factor. For e~a"'ple, al~houg" the r"tn gage "'ould be an indoor
<br />r~cordtn!\ type, a paper roll ...ould need t<> be "eptaced every 30 days.
<br />The degree of tnterest affect. the read1"..as of " volunteer net"'ork.
<br />In areas such as the5us'lllehannal\tver busIn, the freqoencyof floods
<br />h~lp3 to matnt.1in tMere~t in voluntecr warntng net",orks. (Telephone
<br />c;::mve,",'tion ",tth llr. Ste",,,rt K. ';rlgh~, ~lts~lI('h~"n~ Rtv~r B~sin
<br />C01Il;oisston, H O~tob"r 1979). ,n ~he C~cbe ill Poudr. 'tiver hast" are<l,
<br />h""pver, th"r~ have be~<> lo\tg i.'te<~at.. h~t~~"" O'Iajor f..iofall no"ds,
<br />
<br />The elecrroni~ operati<.>ll of ~he automattc ra1\t gages and streatll
<br />gagas is pre.umed to be rdi..ble. St"tionol'eratio\tal conditton lo'Ould
<br />bereportedatleasrda.llY<lndreplaceOlentwouldpreau",ablybelllade.
<br />Automattc gages are SllSCepLible to vanJdlts",. .he automatIc rain gages
<br />"'O\lld be of the tipping bucket ~ype; to reghter rainfall I1c~urately,
<br />they must be tnat<llled sO the bucket device ls level. I r"ln gages
<br />are left out aii yesr, ~n antifreeze solution is needed to ~elt sno", so
<br />lt ca.n be registered on a gage, In ~ m<ljor flood, the "trea~ gages ~y
<br />be destroyed and would ha~e to be replaced. Tile rel1ahiltty <)f gages
<br />is affected by ~tntananc". The usefulness of the ~~g flash fl<>od alar~
<br />
<br />~tf"~'" 8"g:~~,
<br />
<br />",hteh :report by tele~ho""-. has been v<lr1ahle; it depends
<br />
<br />onthedegrecoflocaltnterest"ndgagefllllintenance, (Telephone con-
<br />vers"tlon >ltth M~. Ken King, Nattonal Wc"ther Service River For"~ast
<br />Center, Kans1s City, 7 Dece~h"r 1979).
<br />
<br />102
<br />
<br />c"
<br />
|