Abstract
The results of a three-year randomized seeding experiment carried out as part of the National Hail Research Experiment are described in a nine-part series. In this first part the design of the statistical experiment is presented, including the physical hypothesis for hail suppression by seeding, the seeding procedures, the randomization scheme and the method of evaluation. The experiment was of the singlearea type and the randomization was by day. The criterion for selecting sample days involved the magnitude of radar returns from cumulonimbus clouds over and near the target area. Storms on approximately half the days were seeded according to a random selection procedure. The seeding method involved dispersing silver iodide in the storm updraft from aircraft maneuvering just below cloud base. An air-borne rocket system designed to deliver the nucleant at about the −5°C level in the storm was also employed in one season. The evaluation was based on the measurement of hailfall over a network of instruments in a fixed target area on seed and control days. The total mass of hail falling in the network was chosen as the primary measure of seeding effect. Several other hail and rain measures are also examined and a variety of post hoc analyses are presented in the later papers of this series in an attempt to expose seeding effects if they exist and to give further insight into the results of the primary analysis.
