Estimation of the Effect of Operational Seeding on Rain Amounts in Israel

Abstract

During the period 1961–75, two cloud seeding experiments were carried out in Israel. The first Israeli experiment had a two-target crossover design. The results indicated a positive seeding effect of 15%, significant at 0.9%. The second experiment indicated an enhancement of 13%, significant at 2.8%, in rain amounts in the northern part of Israel. Since November 1975, seeding in the north has been operational. The objective of this paper is to develop a method for estimating the seeding effect of the operational seeding. The proposed method is based on a historical logarithmic model of target precipitation on control precipitation. It is argued that the validity of the proposed approach stems from the stability of the relation between target and control precipitation over time, and the robustness of the meteorological system in the region. The seeding effect in the operational period 1976–90, on the annual rainfall, is estimated by a 6% increase in rain amounts, with a 95% confidence interval of (1.01, 1.12). An analysis of the sensitivity of the effect estimate to the choice of period shows a persistent indication of a positive seeding effect. Possible explanations for the reduction in seeding effect, in comparison to the Israel-2 experiment, are discussed.