Laserfiche WebLink
<br />exploratory rather than confirmatory and covered a limited area because of resource constraints. <br />it may be used as a model for future statistical dcsigns. <br /> <br />Although randomization is the only sure safeguard against bias and its influence on evaluation <br />results, Silverman (Appendix A) sho\...'ed that the use of a bias-adjusted regression ratio in <br />evaluating operational (non.randomizcd) cloud seeding programs in the Sicrra watershcds <br />provided robust results. He attributed this robustness to thc following: (I) the historical and <br />operational periods were quite long, so that the potential efTect on avcragc strcamt10ws from <br />ycar-to-ycar variability and short-tcrm cyclcs was mitigated; (2) the bias-adjusted regression <br />ratio takes into account thc effcct of the long-term trend in natural streamflow through thc <br />regrcssion between the targct and control, and (3) the ratio statistics mcthodology is very robust <br />to departures from its inhcrent assumptions, and was shown to approximate resuhs from re. <br />randomization (permutation) analyses extremely well. <br /> <br />4. Benefits and costs vs other water augmentation technologies <br /> <br />The costs of wcather modification progrJrns are otlen exprcsscd per acre foot (ae.tl) of \\!ater <br />they produce. These estimates depend on thc value of water, which of course varies with local <br />markets and the use to which the water is put. Also, the cost of operational weather modification <br />programs varies with generator configuration, seeding agents. etc. Because demand for water in <br />the Wcst is increasing, so is its value. Agricultural watcr in California is valued from 540 to 550 <br />per ac-ft ($175 per ac.tl during drought), while the average value for hydroc1ectric use (by <br />PG&E) is 5100 per ac.O (Byron ~ar1er, pcrsonal communication), Municipal and industrial <br />values are generally higher, from 5300-600 per ac_ft11. <br /> <br />The Wyoming pilot project conservatively cstimates a weather modification cost bctween $3,96 <br />and $7.91 per ac.tl with associated bcncfit-to-cost ratios of 2.4 to one12. Benelit-to.cost ratios of <br />3:1 to 10:1 were estimated for 10% mountain snowfall increases in thc Sevier River basin in <br />Utah1J, We arc unaware of any calculations of benefits to ski areas. but they are belicvcd to be <br />high since several ski areas havc invested in the technology. Thc Utah Division of Watcr <br />Resources has stated that the estimatcd direct cost of water from an 8 to 12% increase in <br />snowpack from cloud seeding in key mountain watershcds is about $ 1 per ac_tl14, Nevada <br />augmentation estimatcs have Icd to cost estimatcs of S6 to 12 pcr ac-tl. In Colorado, costs for <br />cloud sccding generally would be less than $20 per ac.j), with existing programs costing about <br />one-third that ot'ncv.' programs, This is because much background work has been completed and <br />instrumentation arrays arc already in place. The California DWR~2 has estimated that an <br />additional 300.000.400,000 ac.j) ofncw supply could be rcalized by seeding. with an invcstment <br />of around $7 million. This represents a cost of about S19 per ae-tl, which includes an initial <br />investment of an estimated $ 1.5 to 2 million in planning and environmental studies. These costs <br />do not include randomization or evaluation components. which are recommended additions to <br />ongoing programs. State law mandates that water from cloud seeding is treated the same as <br />natural supply with regard to water rights. <br /> <br />24 <br />