Abstract
Four seasons of hail data were gathered on a randomized cloud seeding project aimed at reducing hail damage and increasing rainfall in western North Dakota. Hail on seed days was generally less severe than on no-seed days. Statistical tests of data from passive hail indicators do not permit rejection of the null hypothesis at the 90% confidence level, but application of rank tests to crop-hail insurance loss data indicates that the seeding reduced crop damage from hail.Post-analyses of related data indicate that 1) the ratio of rainfall amount to hail energy was greater for seed days than no-seed days, and 2) radar characteristics of seeded storms differ from those of unseeded storms. In addition, case studies of 34 storms indicate that damaging hail was usually suppressed when their updraft areas were seeded continuously.
