Laserfiche WebLink
<br />a positive effect is far from conclusive;-however, the indica- <br />tions of the preliminary analysis were that the overall apparent <br />effect of seeding all categories of storms was to increase pre- <br />cjpttation by 16%. When the data was stratified by storm magni- <br />tude, the apparent effect of seeding was to reduce precipitation <br />in light and marginal storms. If marginal storms are excluded, <br />the apparent increase is 19%; if light storms are also <br />excluded, the apparent increase is 25%. The corresponding over- <br />all increases in average annual precipitation were: seeding all <br />storms (A+B+C+O), 3.52 inches; seeding all storms except <br />marginal ones (A+B+C), 4.10 inches; seeding only strong and <br />moderate storms (A+B), 5.10 inches of water. The equivalent <br />"increases in the total volume of precipitation over the analysis <br />area were 169, 197 and 245 thousand acre feet respectively. <br />Analyses using other stratifications may give more conclu- <br />sive results, but possibly the most significant improvement <br />would result from an analysis which took into account the limita- <br />tions which were tmposed on the seeding operations by the need <br />to observe safety restrictions. <br />A conclusive demonstration of an effect due to seeding is <br />rendered difficult by the great variability of storms (in which <br />the mean precipitation ranged from zero to over four inches of <br />water) and by the fact that only a modest number of cases can <br />be studied, the effect of seeding ~eing evidently not so drama- <br />tic that the difference between seeded and not seeded storms <br />stands out clearly. <br />As mentioned above, the. indications of increases in pre- <br />cipitation are limited to strong and moderate storms, numbering <br />90 over the three"seasons. If the true effect of seeding these <br />storms had been the 25%-increase actually indicated, then the <br />chance of detecting this effect, significant at the conventional <br />5% level, would have been only one in about 1 in 5. In order to <br />have an even chance of achieving suc~ess as just defined, it <br />would have been necessary to observe about 250 A and B type <br /> <br />A-3 <br />