Evaluation of Effects of Airborne Silver-Iodide Seeding of Convective Clouds

Abstract

Observations of the convective clouds over the mountain ranges of southeastern Arizona show that a large fraction of them reach temperatures far below freezing but fail to produce precipitation. During the summer periods of the years 1957 through 1960, a randomized cloud seeding experiment was conducted to test if these clouds could be modified by airborne silveriodide seeding. The essential features of the experimental design were the following: 1) An objective technique was evolved for predicting days with suitable clouds; 2) one of a pair of two adjacent days was seeded on a random basis; 3) measurements on all days were made with rain gages, radar and a pair of ground-located aerial cameras mounted at the ends of a 3-mile base leg, and visual observations were made of cloud-to-ground lightning; 4) the signrank test was used to make statistical evaluation of the effects of seeding. The results after four years are the following: 1, The rainfall statistics do not allow a conclusion that seeding had any effect. Statistically, the rainfall on both seeded and not-seeded days was the same. However, the sensitivity of the tests to changes in the quantity of precipitation was quite low. 2. The frequency of occasions of large thunderstorm echoes was statistically the same on seeded and not-seeded days. 3. The cloud-census and radar studies suggest that in some cases, seeding may have led to the initiation of precipitation echoes. 4. Although there were more cloud-to-ground lightning strokes on seeded days, the differences between their occurrence on seeded and not-seeded days were not statistically significant. Evidence is presented indicating that precipitation in convective clouds is not initiated by the ice-crystal process. Samples of days with heavy rain and light rain were compared in order to study the factors governing the quantity of rain. It was not possible to show that there was any relation between the region of echo initiation and the quantity of precipitation. On the other hand, it is clear that on days with heavy rain there were many more large clouds. The observations are best explained if it is assumed that the quantity of rain is governed by those properties of the atmosphere which determine the number and size of the convective clouds. The microphysical processes which determine the region of precipitation initiation do not appear to be as important as was once suspected. A new program of seeding tests is described which should be more sensitive to changes in quantity of rainfall. It differs in several fundamental aspects from the program conducted during the period 1957 to 1960.