|
<br />323
<br />
<br />\
<br />
<br />JOURNAL OF APPLIED METEOROLOGY
<br />
<br />VOLUME 12
<br />
<br />
<br />~
<br />
<br />. Run Otl ')'0''')1\\ It'! ..!~~l_ Ufll,,.,, In Th" C;'ud,
<br />
<br />P'fo('p'lal,on SloltOr,\ U...d ro' Mult,ple COlre-kJlIt," W,th Run Off
<br />StatIon' In Ta'ge'
<br />
<br />--- Boundar, Of Tarqel A"o o.t",.O In '",' S'ud,
<br />
<br />_ BwndQty 01 lOfl~.' At.u 0.'''1" b, l". 6\)I.ou Q1 ~\orI\ohQl'o.
<br />.CoIorodo R...., RoSIn P,lot ProJect" u.p 01 In"tlor ~p"'.'
<br />
<br />@ e,l,
<br />
<br />'-...,
<br />f,.._.:~~;~j~.
<br />
<br />Q
<br />
<br />" ~ r"_,
<br />1)112101 ,} .. --,
<br />
<br />, ' ""'~, '---,
<br />
<br />,..
<br />
<br />~.~"','
<br />'~
<br />1"";01('(
<br />
<br />~lf?t.J"<lt
<br />
<br />--1
<br />
<br />'1>"" ""'I"~'
<br />.._~---
<br />\ ~'" . \ ~,..~
<br />
<br />\~v~oo
<br />1l\
<br />\ f~ \"
<br />\< \~
<br />
<br />1~C'/J"8J~~""
<br />~~
<br />_~ ~r--"
<br />C.....,., -~
<br />
<br />FIG, 1. Target area (San Juan Mountains of Colorado) showing zones 1, 2, 3 and 4, runoff stations,
<br />and precipitation stations,
<br />
<br />exceeds the expected range of increases induced by
<br />man, Simple statistical tests have been developed
<br />(Markovic, 1966). They have not proven very sensitive
<br />and, as a result, require long periods of observations,
<br />prior to and during seeding operations, in order to
<br />give satisfactory test results. Furthermore, these tests
<br />are insensitive when experiments are performed during
<br />a dry period of annual stream-flow sequences. In the
<br />early days of weather modification it was thought that
<br />such simple techniques would be adequate because it
<br />was hoped that the increases would be large. By this
<br />simple technique (two-sample test) it can be shown
<br />that the number of years LV needed for significance at the
<br />95% level (two-tailed) and 50% power (Brownlee,
<br />
<br />1961) is
<br />
<br />LV = (1.96)2Cv2/h2~4Cv2/h2,
<br />
<br />(1)
<br />
<br />where Cv is the coefficient of variation, i.e., the percent
<br />ratio of standard deviation to the mean, and h is a given
<br />(hypothesized) percentage increase in the population
<br />mean. The number LV was calculated for a variety of
<br />basins in the target area under the assumption of a 10%
<br />increase in "seasonal" runoff (total runoff for the
<br />6-month period March-August inclusive). The results
<br />are shown in Table 1. These numbers greatly exceed
<br />the 4-5 years planned duration. Even for a 201/,0
<br />increase the numbers would be prohibitively high. (The
<br />actual number of years would be the ones given in the
<br />
|