Development and Application of a Predicator Control for the Evaluation of a Winter Orographic Cloud Seeding Project

Evaluation of an operational-type winter cloud seeding project in Utah is made by developing meteorological predictors of target precipitation. Predictors (covariates) are developed by matching 1200 GMT rawinsonde data and 24 h precipitation amounts. These predictors and precipitations are summed over seven unseeded seasons to form a seasonal predictor-predictand relationship, for which the correlation is 0.975 when the average precipitation for all stations is used, and 0.879 when only the two highest altitude stations are used. Then, the predictor is found for each of the seeded seasons, and based upon the unseeded predictor-predictand relationship, the predicted precipitation is obtained. Differences between predicted and observed precipitation in seeded years are compared and tested for seeding effects. Application of the method to the first two years of the project indicates a substantial chance that little or no effect of seeding occurred. It is concluded that the method offers a promising approach to the evaluation of winter cloud seeding projects.