TL;DR: The North Dakota Cloud Modification Project (NDCMP) as mentioned in this paper uses an acetone-based AgI solution, end-burning AgI flares, ejectable AgI flare, and dry ice pellets (CO2) as seeding agents.
Abstract: Operational cloud seeding has been conducted over parts of western North Dakota since the 1950’s. Operations began with ground based seeding in the early years, but by the early 1960’s airborne delivery of seeding agents became the preferred method. Currently, a combination of cloud base and cloud top seeding is employed on the North Dakota Cloud Modification Project (NDCMP) using an acetone-based AgI solution, end-burning AgI flares, ejectable AgI flares, and dry ice pellets (CO2) as seeding agents. The goals of the dual-purpose project are to increase growing season precipitation and reduce crop and property damage caused by hail. The project is funded by county tax levies and limited state cost-sharing. Operations are conducted from June 1 through August 31 each year, twenty-four hours a day, seven days a week. Every evaluation of the project has indicated beneficial results, but with varying statistical confidence. Rainfall increases on the order of 7 to 15% have been realized in and slightly downwind of the target counties. Crop-hall damage as been reduced by 45% and wheat yields have increased by 5.9%. A benefit-to-cost ratio of 35:1 has been realized for wheat production alone.
TL;DR: In this article, a bulk-water microphysical convective cloud model is used to investigate the maximum effect of hygroscopic seeding on small to moderate size clouds but its effect is much less definite on larger clouds.
Abstract: A bulk-water microphysical convective cloud model is used to investigate the maximum effect of hygroscopic seeding. Using the fact that the desired cloud seeding effect is to transform the cloud from a continental one to a maritime one makes hygroscopic seeding a particularly easy process to simulate in bulk-water microphysics cloud models. In those types of cloud models, one only has to change the autoconversion threshold (i.e., the onset of coalescence) and the number concentration and dispersion characteristics of the cloud water to switch from a continental-type cloud to a maritime-type cloud. The continental cloud has no autoconversion, the maritime cloud an extremely efficient autoconversion process. Admittedly, this tests two extreme situations and makes the assumption that the effects of hygroscopic seeding will be somewhere between the two extremes, the maximum effect being represented by the maritime-type cloud. Very preliminary results indicate that the hygroscopic seeding would be quite efficient on small to moderate size clouds but its effect is much less definite on larger clouds. The dynamics of the clouds are crucial to the results.