Seeding Cumulus in Florida: New 1970 Results

See allHide authors and affiliations

Science  09 Apr 1971:
Vol. 172, Issue 3979, pp. 117-126
DOI: 10.1126/science.172.3979.117


In the Florida single cloud experiments, the main result of the statistical analyses is that the dynamic seeding effect on rainfall is large, positive, and significant. From all the 1968 and 1970 data together, the seeding effect is estimated to be larger than a factor of 3; that is, the seeded clouds rained more than three times as much as the controls after the seeding run. On fair days, defined objectively by percentage of area covered by showers, the seeding effect is shown to be larger than the overall average, but it may be negative on rainy days. Rainy days in the tropics are about 10 percent of the days with rain, but they produce about half the total rainfall. The applicability of our single cloud results to other areas is not established but seems hopeful for many tropical and subtropical regions. It can be assessed by cloud population studies together with our numerical model (25).

Guidance for the next steps toward practical rainfall enhancement and toward the understanding and modification of cloud systems in storms may be provided by our study of merger clouds. Mergers are shown often to produce more than an order of magnitude more rain than isolated clouds on the same day, probably owing to dynamic invigoration of the merged cloud circulations. Results of our first small attempt toward inducing and documenting mergers in a multiple cloud seeding experiment appear promising. Although far from statistically conclusive, they have opened a new frontier in the science and technology of dynamic cloud modification. It is also hoped that the multiple cumulus seeding experiments will help to clarify the formation of "cloud clusters" and their role in large-scale circulations, thus contributing to the focal subject of the Global Atmospheric Research Program in the tropics.