WMOD00464 - Laserfiche WebLink

Home Browse Search

B-10 .075 is at its 27% point .135 is at its 11 % point .425 is at its 21% point .89 is at its 63% point With allowance for multiplicity, then, there is no alternative to declaring that this experiment has not reached a positive conclusion. Repetition focused on Autumn, or possibly on Autumn and Winter, would seem most likely to lead to confirmation. 11. Overall evaluatio~ of experiments In this section, we present summary comments on a selected set of recently completed weather modification projects. We note, particularly, conclusions (if any) that the prudent per- son might draw from each. Drs. Brier, Mielke and Biondini prepared a "check-list" description of each experiment, following the outline of Section 46. Our comments are based on those descriptions (which we hope will be made available somehow) and on some of the published reports. We have NOT, with minor exceptions, reanalyzed any data. * FACE * FACE was an exploratory experiment conducted over a substantial area of South Florida during the summers of 1970-1976 (except for 1972 and 1974), to assess the effects of dynamic seeding of convective systems. In August, 1975, and thereafter, a new flare type was used for seeding, one that had been selected from laboratory experiments to be more effective. Over all years and all conditions, results show a slight increase in mean rainfall on days subject to dynamic seeding, but the observed increase fails to approach statistical significance. When data are separated by year, it is found that for 1970-73 (24 experimental days), average rainfall for seeded days was slightly less than for nons.eeded days. For 1975 and 1976 (74 experimental days) average rainfall on seeded days was found to be greater (by some 30 percent) than for nonseeded days. However, the observed increase in average rainfall under seeding conditions in 1975 and 1976 is largely attributable to only a few (5 or 6) days. These later years were years during which the presumably more effective flares were employed. A number of other separations were employed in an effort to achieve greater sensitivity. The numerous analyses that stem from various separations of data give rise to problems of multiplicity. It is noteworthy, however, that those portions of separations with strong physical bases -- light wind (vs. strong wind), undisturbed (vs. disturbed) days, days with echo motion (vs. no echo motion), large (vs. small) expenditure of flares -- tend to be associated with larger increases in mean rainfall under the seeded condition. In view of the limited sample size, there is little hope of statistical significance here once multiplicity is allowed for. Results suggest the possibility of a seeding effect that, under some conditions, increases rainfall. At this point, due to problems of multiplicity, (expected in an exploratory phaseD and to the possibility of subjective influences of knowledge of seeding, conclusions remain uncer- tain. .../ * NHRE * Results from this study are inconclusive. Contrary to initial hypotheses, the mean hail mass on seeded days was found to exceed that on nonseeded days, and the difference approaches statistical significance if the few (6) observations from 1973 are excluded.