Laserfiche WebLink
<br />Occasionally. samplcs of newly fallen snow are collected for an analysis of silver contenl. <br />This is an evaluation technique encountered more frcquently in research projects due to the <br />expcnse involved. Snow samples collected prior to cloud seeding or from non-seeded stonns are <br />analyzed to establish the natural background silver content (if mcasurable \vith available analysis <br />techniques) for comparison with snow samples taken from seeded stonns. This technique is only <br />valid for projects using silvcr iodide as thc cloud seeding agent. although some analysis techniqucs <br />are applicablc to other possible cloud seeding agents as wcll (i.c.. lead iodide). Several analysis <br />tcchniqucs have been developed for use in such analyses. including neutron activation. proton <br />excitation. and t1ameless atomic absorption. An examplc of an analysis of the downwind transport <br />of silver iodide outside of primary target arcas is given by (Warburton 1974). Warburton et at <br />1996 demonstrates how tracc chemical assessment techniques strcngthen traditional target and <br />control precipitation analyses. <br /> <br />^ modification of this trace chcmistry assessment technique involves the simultancous <br />releasc ofa control aerosol along with an active secding acrosol (Warburton ct al. 1995). Such <br />tracers have properties very similar to thc sceding agent. with the key exception that they do not <br />nucleate ice. Insolublc in water, they have an extremely low natural background in precipitation <br />and arc only removed from the atmosphere by passive precipitation scavenging mechanisms. <br />Both the seeding agent and traecr arc transported and scavenged in a very similar manner when <br />conditions are not conducive for effective seeding. Givcn similar release rates. detecting the <br />same concentrations of silvcr and thc tracer. c.g.. indium. in precipitation samples at dowmvind <br />locations indicates that thc t\VO acrosols were most likely rcmoved from thc atmosphere solely by <br />scavenging. On the other hand. \I.'hen sufficient supercooled liquid water (SLW) e.xists and <br />tcmperatures arc cold enough for the active seeding material to nucleate new icc crystals. the <br />ratio of silvcr to traccr in target area precipitation samplcs can be much greatcr than unity. This <br />indicates that some fraction of the secding material was directly responsible for the nuclcation of <br />ice cr~'stals that evcntually produced additional snowfall. <br /> <br />13.3 l\lodclin2 Annroaches <br /> <br />Sophisticated atmospheric computer modcls have the potential to calculate the amounts <br />of natural precipitation for short intervals (e.g.. 6 hours. 12 hours) in mountainous areas. !fthese <br />predictions arc validated as accuratc.thcy could be comparcd with the amount of precipitation <br />that fell during seeded periods within the intended target area to detennine the impact of seeding <br />on target area precipitation. An attempt to vcrify the output of the RAf\..IS computer model <br />developed at Colorado State Univcrsity versus observed and predicted modilicd precipitation uue <br />to cloud seeding was madc for the 2003-200-l \vinter season in ccntral Colorado. with rather <br />mixed results. This work was done under the Colorado WDMP. Some of the conclusions from <br />the tinal report (Colorado Water Conservation Board. 2005) are: <br /> <br />. When model simulated precipitation was compared to measured 24 hour <br />prccipitation at 61 SNOTEL sites the model exhibitcd a mean precipitation bias of <br />1.88. <br />