Statistical evaluation of a cloud seeding experiment in coahuila, mexico

During the summers of 1997 and 1998, the PARC hygroscopic cloud seeding experiment was conducted in Coahuila, Mexico. This randomized experiment yielded 94 cases, 43 seeded and 51 unseeded. Each storm was tracked objectively, and various characteristics were measured at 5-minute intervals out to 60 minutes after seeding. Primary measures of the effects of seeding include rain mass, which is the focus of this analysis. Exploratory Data Analysis (EDA) techniques are used to display and compare the two groups of storms. The results of one-tailed hypothesis tests are also presented. Permutation tests are used to compare the differences in the medians, quartiles, and means of both samples. Preliminary analyses suggest that cloud seeding had a positive effect on the amount of rain produced by thunderstorms. Further, time series analyses of rain mass suggest that the seeding effect was due to an increase in the life span of the seeded storms rather than an increase in the precipitation rate. Statistical issues associated with this evaluation are also considered.