A SURVEY AND ASSESSMENT OF THE DESIGN AND EVALUATION TECHNIQUES OF CUMULUS CLOUDMODIFICATION EXPERIMENTS FOR RAIN ENHANCEMENT.

Abstract

Physical basis of cloud seeding and meteorological considerations involved in the efficient design of experiments for rain enhancement are discussed. An analysis of thevarious designs and statistical techniques currently being employed for evaluating the experiments is presented. The limitations of available statistical methods, when used alone, to establish significantly the changes in rainfall caused by cloud seeding within a reasonable period of experimentation are broughtout. The manner in which remarkable increase is produced in the power efficiency of some tests when applied along with the
simultaneous measurements of physical covariates or predictor variables and proper stratification of data is elucidated. The scope of anlayses based on postfactum stratifications or partitioning of data and the problems of multiplicity of analyses are discussed.