|
<br />:.~#.;"j.":;.:".<;.:C.L '.."..b';~i '.';;';:'''''';'':';'''il;~;,,",:..i ...",.>.:i'''>im;""~";":';;;''''"''~'';ii?"",.';''';';.",,,,,,,,:,~;,;~~:.. "f.''' "'''.''"~4'''-'''''.'''_'--
<br />
<br />:]9
<br />
<br />~)
<br />
<br />!ai
<br />
<br />5)}
<br />
<br />ia~
<br />fi-.
<br />
<br />.'
<br />;.&.~
<br />
<br />~
<br />~:
<br />if..
<br />ll)
<br />:IS
<br />i;
<br />e
<br />:a
<br />it
<br />F
<br />E!-
<br />ll,
<br />
<br />~
<br />
<br />cf
<br />~.
<br />
<br />-
<br />
<br />OcroBER1980
<br />
<br />JAMES A. HEIMBACH, JR., AND ARLlN II. SUPER
<br />
<br />1179
<br />
<br />location is an oversimplification, especially as only
<br />a single climatological Z-R relationship was used.
<br />This may increase or decrease the simulated sam-
<br />pling and among-storms variances. While this effect
<br />is believed to be limited, there is no known way of
<br />evaluating it with the available data set.
<br />Fig. 2 depicts the details of a particular CC sam-
<br />pled by the various networks used with the gage
<br />spacing and total rainfall accumulations noted. The
<br />CC was missed entirely by the 2771 km2 per gage
<br />network and nothing more than a trace was detected
<br />for 1386 km2 per gage. Although definition is lost,
<br />the total accumulation estimate remains reasonably
<br />stable through about 346 km2 per gage. This example
<br />was classed as a small storm.
<br />
<br />
<br />10.8km2/QOge
<br />2.0X105m3
<br />
<br />d. Characteristics of simulated CC samples
<br />
<br />To estimate the impact of storm size, the 103 CC
<br />sample was divided into two halves based on the
<br />"ground truth" median of 2.47 x 105 m3 precipita-
<br />tion accumulation. Consideration by size (total rain
<br />volume) should suggest how much the experimental
<br />duration might be reduced if an accurate fore-
<br />casting or nowcasting ability was developed.
<br />The mean rainfall accumulations y were rather
<br />consistant through the denser gage spacings for the
<br />smaller half, the larger half, and all 103 CC's com-
<br />bined (see Fig. 3). The reductions from "ground
<br />truth" to 10.8 km2 per gage resulted from missing
<br />a few extreme range bin values. The variances S2
<br />
<br />tt///--;: 21.7 km~~e
<br />,/ ..'/ 2.0:<10 m
<br />.' . :
<br />" . :
<br />, - ,-_1#
<br />, / .
<br />I ,
<br />~.....,,,
<br />
<br />'.
<br />
<br />~"".--': 86.ISkm2/QoQe
<br />,/ /:' 2.4'<105m3
<br />, '
<br />, .
<br />, .
<br />.' ... --_..,' . .
<br />\ ,I'"
<br />1Io......
<br />
<br />,"'o-~"
<br />.... .,,''' 346km2,/oOQ8
<br />. . ,..", U~XI0!Sm3
<br />:0' "...'
<br />~... .,;'
<br />
<br />/'--\ 1386km2,IQo;e
<br />, .
<br />,/ J nothIng above
<br />: / trac:.
<br />. ,
<br />. ,
<br />. ,
<br />. ,
<br />. ,
<br />..........
<br />
<br />LEGEND
<br />... !SOkm >1
<br />________ Trace preClpltotion (< 0.01 inch)
<br />C :t 0.01 inch
<br /><!) ) 0.02 inch
<br />4& ~O.03jnch
<br />
<br />FIG. 2. Simulated sampling of t CC for various raingag.e densities.
<br />./
<br />
<br />~. "'~~\~~~JP;-4.~",~"'''"~,,,!~x~';~''1\f7~.~.1''/~r~':;.,?:,~r?~'~5~~~,.~jot~',::;;.~r-.-;~ r.~'fr~.r-". !.r-,r.::...~~~T'r.Ti~ ,~;.~~ ~~?~::J;' ::'~~:S:42~~~:.:r.;''--''~''I!l''~::'<~"'!''.:'~~'.~r~.~~~~~'''''
<br />
<br />,..,
<br />
<br />!?J/..' 'j 43.3 km~gage
<br />:' .... 2.0XI05ro3
<br />, ,
<br />. .
<br />, .
<br />, .
<br />, .
<br />: ........'
<br />~............
<br />
<br />.-....
<br />
<br />If../ ) 17'3km~og8
<br />/ /' 2.2X105m3
<br />. .'
<br />. "
<br />..-......
<br />
<br />--- 69'3km2/QoQ8
<br />('0'\ 1.6X 105m3
<br />\ .,,'.
<br />............'
<br />
|