Microphysical and Radar Observations of Seeded and Nonseeded Continental Cumulus Clouds

Abstract

Controlled cloud seeding experiments were conducted near Bethlehem, South Africa during the summer of 1984?85. The experimental unit was the semi-isolated cumulus congestus cloud. Microphysical measurements were obtained by three instrumented aircraft flying in stacked formation. Radar observations were made by a 5-cm wavelength radar performing volume scans at 5-min intervals. A three-way balanced randomization scheme was used to seed the clouds near the ?8°C level with either dry ice pellets, silver iodide pyrotechnics, or a placebo (no-seed) treatment. Strict cloud selection criteria, based on the measurements made during an initial inspection penetration, assured the treatment of clouds in their developing stages as their tops row up through the ?10°C level. A total of 60 clouds were chosen and treated. Using ice crystal measurements, mostly above the seeding level, it was conclusively demonstrated that some cumulus clouds were modified by the injection of either dry ice or silver iodide. High ice concentrations were produced and the evolution of the ice particle size distribution is consistent with a rain enhancement hypothesis involving an artificially induced ice embryo process. However, the liquid water contents decayed rapidly, primarily due to entrainment, and in the majority of the seeded cases precipitation particles formed due to low-density accretion onto aggregates resulting in little or no precipitation at the ground. The dry ice and silver iodide seeded clouds that echoed had significantly higher maximum 1-km average ice concentrations than the placebo clouds, as measured by the research aircraft during the clouds' developing stages. Approximately 67% of the dry ice-treated clouds, 63% of the silver iodide?treated clouds and 45% of the placebo clouds produced radar echoes >10 dBZ. The dry ice-seeded clouds that echoed had significantly higher maximum cell heights than the placebo cells and the silver iodide-seeded cells. When the test cloud data were stratified according to cloud diameter, a positive association existed between seeding and the number of clouds that echoed. Wider clouds responded more favorably to seeding, presumably because they were affected less by entrainment.