|
<br />
<br />the experimental area. A limited number of physical
<br />case study experiments. which released either
<br />propane or Agl from lIAS. showed obvious seeding
<br />elfects at TAR.
<br />
<br />An icing rate sensor just upwind of the propane
<br />dispensers wa.~ LL..ro to initiate a 2 hour experimental
<br />block whenever SL W eloud was observed to exceed
<br />a threshold. A programmed data logger then made a
<br />
<br />In,,,",,, "'-;>0;"''''''
<br />
<br />",.;>ovw
<br />
<br />,.,~....
<br />
<br />"'a1U'''
<br />
<br />nItT...~""'''''"Obf--
<br />rodJD.(;..._
<br />....$~......_..
<br />.n<l....._ot>o
<br />15Ok_00..
<br />SR....._Clh
<br />G:sC T.lI"I
<br />GfUl T.....
<br />GTltT.....
<br />COOlT.ll'"
<br />G:sOT."",
<br />~,-
<br />
<br />J$"311l">
<br />
<br />..JI-......
<br />
<br />JlI'nlr'>,
<br />
<br />random decision to seed either the first or second of
<br />the two -1.0 minute experimental units (EUs) within
<br />the 2 hour block, with the other unit recei\"ing no
<br />seeding. The last twenty minUles of each hour WeTe
<br />used as a bulTer period. to minimize seeding
<br />contamination of any following nonseeded EU. The
<br />population of non seeded units (placebos) provided
<br />the basis of comparison for statistical tesling of
<br />propane seeding effectiveness in increasing snowfall.
<br />
<br />"
<br />
<br />n'""","",
<br />
<br />"..,.."'.....
<br />
<br />"'.,BV'.... '''''''-1II.
<br />
<br />~..,,'.-
<br />
<br />~f~
<br />y ! ~~
<br />~
<br />
<br />..'<n:'"
<br />
<br />JlI',O'I)"'<
<br />
<br />""'J"lO"~
<br />
<br />~
<br />
<br />n'",,,,,,,,, ",-,,-:arv. ''''.'1TW ''''U:I'''''' n...,'........ '''''.",,,,, ,....~lI'W "",nJ"Y,' ",.,~'7V>
<br />
<br />':.._ . /._., _I
<br />.--.......- ,-
<br />
<br />Figure I. \lap oflhe expcnmenlal area and site locations for the Lilah 2003-2()()-1. r.mdomized propane doud seeding
<br />expenml'Tll. The conlour inlen;al i~ 100 meters.
<br />
<br />Experimenta] unit snow wate:r equivalent (S\\'E)
<br />obse:r.3Iions provide:d the: test variable f(lr e:ach EU.
<br />These data were obtained by htgh resolUlion digitized
<br />precipitation gauges. Time:s of e:stimated seeding
<br />plume (or placebo) passage over the TAR and each
<br />gauge were: based on liAS and TAR wind speeds and
<br />a relationship developed from prior plume tracing
<br />experiments. Fi\e gauges were located on the
<br />windward slope and plateau top along the expected
<br />plume trajectory. This nern-ork ~nniltc:d
<br />investigation of seeding elTectiveness by distance
<br />from about 2 10 6.5 km do....nwind of the liAS. One:
<br />of the gauge locations, GRD. proved to ~ too windy
<br />for suitable observalions as dis.;usse:d in the [\\-'0
<br />
<br />references. and will not be considered funher.
<br />Control gauge GNO was also maintained on the
<br />plateau top. crosswind of the seeding plume. II was
<br />well-correlated with the remaining fOllr target gauges.
<br />\leasurements at G1'\O were essential t(l reducing the
<br />high natural variance (variability) in snowfall during
<br />statistical testing. Wind. temper.Uure. icing which
<br />indicates SL \1': presence, and other obsen'ations were
<br />made at the "AS and TAR for the: purpose of
<br />panitioning the precipitalion data set into
<br />subpopu]ations for most of the testing. Some limited
<br />observations were made at stations SIR and TAD
<br />shown on Fig. l. bUI the:y had minima] impact on the
<br />statistical testing. For all practical purposes the map
<br />
<br />]
<br />
|