RESPONSES OF WARM-BASED, MIDWESTERN CUMULUS CONGESTUS TO DYNAMIC SEEDING TRIALS

Findings are reported from an analysis of AgI seeding effects on indiv idual cumulus congestus clouds in the 1989 Illinois Exploratory Cloud Seeding Experiment. The experiment was designed around a dynamic seedi ng hypothesis. Randomized treatments of individual clouds were based o n ''floating'' experimental units, initially centered on the first tre ated cloud. The analysis was based on 12 experimental units having a t otal of 67 treated echo cores-32 treated with sand and 35 with AgI. Pr ior to any analysis for seeding effects, a check of many of the physic al conditions at the time of treatment that would govern future cloud growth showed a bias for the sand-treated clouds to be expected to ult imately grow larger than the AgI-treated clouds. Thus, even though ran domization produced numerical balance, direct comparison between the p osttreatment behavior of the entire sample of sand- and AgI-treated ec hoes could not be expected to provide a true impression of possible se eding effects. In an attempt to overcome the bias, an empirically defi ned seedability index composed of criteria consistent with the Illinoi s dynamic seeding hypothesis was developed and applied as a filter to reduce the sample bias, and thereby reveal possible seeding effects. R esults of two representative applications of the seedability index are reported: one for a subgroup of clouds with higher index values, and the other for a subgroup with lower index values. The primary impressi on from the seedability index analysis was that AgI treatment did not have a pronounced initial effect on the behavior of individual echo co res, and that if seeding had any effect at all it may have been negati ve on maximum cloud-top height. This finding was not consistent with t hat expected from the Illinois dynamic seeding hypothesis.