An independent evaluation of a South African hygroscopic cloud seeding experiment, 1991-1995

An independent statistical evaluation of a randomised hygroscopic cloud seeding experiment carried out in South Africa from 1991-1995 is presented, based on the original radar data for each storm. It is found that seeded storms clearly lasted longer than unseeded ones. It is suggested that the initiation of precipitation at a lower height in seeded than in unseeded clouds (demonstrated through microphysical measurements by the experimenters) led to a reduced dispersion by windshear of the precipitation falling from the updraft in the seeded clouds. A more concentrated downdraft region resulted, closer to the updraft. The surface gust front was thereby intensified and its interaction with the storm inflow or other gust fronts enhanced convection. Some consequences of such a mechanism are explored. It is concluded that this experiment offers a real hope of increasing precipitation from convective storms and represents an important advance in weather modification.