First Cloud Seeding Heralds Weather Modification

Vincent Schaefer made a flight near Schenectady, New York, to perform the first scientific seeding of a supercooled cloud with dry ice, marking the first case of weather modification.

Summary of Event

On November 13, 1946, Vincent Joseph Schaefer made the first cloud-seeding flight east of Schenectady, New York. Curtis Talbot Talbot, Curtis piloted the small plane, while Schaefer scattered three pounds of dry ice into a three-mile line of supercooled stratus clouds. Cloud drops turned to ice crystals within five minutes after the dry ice was scattered. Snow formed from the ice crystals fell about two thousand feet below the cloud bank before succumbing to the heat of friction and evaporating. This first cloud-seeding experiment marked the beginning of scientific weather modification. [kw]First Cloud Seeding Heralds Weather Modification (Nov. 13, 1946)
[kw]Cloud Seeding Heralds Weather Modification, First (Nov. 13, 1946)
[kw]Weather Modification, First Cloud Seeding Heralds (Nov. 13, 1946)
[kw]Modification, First Cloud Seeding Heralds Weather (Nov. 13, 1946)
Weather modification
Cloud seeding
Weather modification
Cloud seeding
[g]North America;Nov. 13, 1946: First Cloud Seeding Heralds Weather Modification[01880]
[g]United States;Nov. 13, 1946: First Cloud Seeding Heralds Weather Modification[01880]
[c]Science and technology;Nov. 13, 1946: First Cloud Seeding Heralds Weather Modification[01880]
[c]Earth science;Nov. 13, 1946: First Cloud Seeding Heralds Weather Modification[01880]
[c]Environmental issues;Nov. 13, 1946: First Cloud Seeding Heralds Weather Modification[01880]
Schaefer, Vincent Joseph
Langmuir, Irving
Vonnegut, Bernard

Planes seed clouds over the Pacific Ocean near Sydney, Australia, in 1957.

(National Archives of Australia)

Schaefer and his mentor, Irving Langmuir, had been closely involved with the icing research; initially, Schaefer had been Langmuir’s laboratory assistant. Schaefer’s and Langmuir’s investigations into the formation of ice crystals began in February, 1942, when Langmuir was commissioned by the U.S. Army to work on the development of a generator to produce smoke screens. The goal was to establish the nozzle size and proper pressure of oil vapor needed to generate the desired size of smoke particles. Schaefer and Langmuir found that boiling oil with a pressure of five to fifteen pounds per square inch (psi) produced a jet of vapor that drew in large quantities of air, thereby almost instantly lowering the temperature so that the oil particles had no appreciable vapor pressure. The optimum nozzle size could be calculated from the vapor pressure of the oil, the dilution rate of oil in air, and the rate of temperature decrease.

The next series of experiments leading to cloud seeding was conducted by Langmuir with Schaefer’s assistance between 1943 and 1945. Working for the U.S. Army, Langmuir and Schaefer studied the formation of rime (ice) on aircraft using a set of slowly rotating cylinders of various sizes. They analyzed data on the rate of ice accumulation on the cylinders under a known wind velocity. This led to a method for determining the diameter of water droplets in clouds.

In his evaporation-condensation theory, Langmuir assumed that an excess of condensation nuclei was available for cloud formation. When he applied that theory, he discovered that the number of nuclei initially present in the air had little to do with the number of cloud drops present in any cloud. Instead, the number of droplets in a cloud was dependent on the rate at which air rose into the cloud base.

Langmuir and Schaefer postulated that ice-crystal nuclei in clouds would be extremely rare. In clouds below the cirrus level, ice crystals in clouds, even at temperatures of -20 degrees Celsius, number no more than 10-9 crystals per cubic centimeter. Supercooled water droplets, on the other hand, exist at concentrations of between 100 and 1,000 crystals per cubic centimeter, yet snow can fall from clouds at temperatures no lower than -5 degrees centigrade. Clearly, factors other than temperature must affect the formation of ice crystals. Two questions occurred to the scientists: Which ice nuclei lead to snow formation? and How could nuclei be introduced into clouds to produce snow?

During the winter of 1945-1946, Langmuir and Schaefer established several cloud characteristics that enabled them to differentiate between water clouds and ice-crystal clouds. They found that when viewed at critical angles, clouds containing snow produced coronas around the sun; when the prisms were pointed downward along their long axis, so-called sundogs became visible. When seen from above, ice-crystal clouds produced a bright spot at a distance below the horizon equal to the height of the sun above the horizon. When a plane flew through a water-drop cloud, droplets produced rime on the plane’s wings, an effect not produced by an ice-crystal cloud.

Langmuir and Schaefer also discovered characteristics enabling differentiation between large and small cloud drops or ice crystals. From above, light intensity reflected from water drops is very intense at angles of 20 to 30 degrees from the sun. From this point to about 138 degrees, light intensity decreases; at 138 degrees, it rises again. With large drops present, a rainbow would be seen at 138 degrees. A so-called glory seen at 138 degrees indicated that the diameter of the water drops was very small compared to the wavelength of light.

When the sun is viewed through a thin water cloud, the edge of the sun’s disk is always sharp, no matter how small the drops. Experiments with smoke particles led to the conclusion that there is no sharp maximum in intensity of light scattered by diffraction through small angles, hence the disk edge is sharp.

Conversely, when the sun is viewed through an ice-crystal cloud sufficiently thick to protect the eyes, the sun disk has a fuzzy edge. When the sun’s intensity is reduced by a factor of ten thousand, the sun’s disk becomes fuzzy. This effect occurs only when the surfaces of reflection are very large compared to the wavelength of light. Hence, large cloud drops and snow crystals produce a fuzzy effect, whereas smaller drops or crystals never do.

The culminating factor that led to Schaefer’s attempt at cloud seeding came from his work with dry ice during the spring of 1946. Using a four-cubic-foot illuminated freezer lined with black velvet, he found that when he breathed into the unit, moisture in his breath condensed and formed cloudlike particles. Even at a temperature of -3 degrees Celsius, no ice particles formed. Schaefer experimented unsuccessfully with a variety of substances. Finally, wanting to lower the temperature, he inserted dry ice, whereupon the air in the freezer instantaneously filled with ice crystals. Later, Schaefer discovered that even small pieces of dry ice, or a needle filled with liquid air pulled through the chamber, could accomplish the same results.

Further experimentation led to the discovery that the dry ice itself did not directly cause formation of ice crystals, but lowering the temperature did. Schaefer eventually determined that a temperature of -39 degrees Celsius was required for the spontaneous formation of ice crystals. In an article published in Science, he outlined his experiments and announced his intent to drop dry-ice particles from an aircraft into clouds to try to change supercooled clouds to ice-crystal clouds on a large scale.

During August of 1946, Langmuir did a theoretical study of the increase of nuclei resulting from the dropping of solid carbon dioxide pellets through supercooled water clouds. He concluded that an ice nucleus can have a diameter of no more than 10-6 centimeters, and the number of nuclei formed by a single pellet could be 1016. If each nucleus could be made to grow to a weight of 10-5 grams, it would be possible theoretically to produce 100,000 tons of snow from 100 cubic kilometers of cloud. Langmuir further calculated that as supercooled water drops evaporate and are deposited on ice crystals as ice, the quantity of frozen water will be greater than the original quantity of liquid water, because vapor pressure is less over crystals than over water drops. Ice crystals grow at the expense of water vapor content in the cloud. Langmuir determined that there are two sources of heat to warm air in a cloud: the heat of fusion and the heat of deposition produced in the conversion of vapor to ice.

Langmuir postulated that warmed air in a cloud causes an upward acceleration, which leads to turbulence. That turbulence in turn brings larger masses of air into the circulation, ice nuclei will be carried away from the seeding plane, and upward velocities will gradually increase. These vertical air currents then carry the nuclei upward and spread them laterally. Nuclei will filter downward, leading to more rapid spreading. In November, 1946, Langmuir calculated that seeding a stratus cloud needed to be done only in lines one to two miles apart, and that complete nucleation ought to occur in about thirty minutes. With these research results in place, Schaefer was prepared to make his historic flight on November 13.

Another scientist, Bernard Vonnegut, who worked in the General Electric laboratory and experimented with lead iodide and silver iodide, concurrently discovered that silver iodide also worked well as a seeding agent. Schaefer contributed some assistance and equipment to Vonnegut. On November 14, 1946, Vonnegut discovered that the smoke of silver iodide produces a good nucleus that is effective at -5 degrees Celsius and below.

In February, 1947, the U.S. Signal Corps began cloud-seeding with experiments named Project Cirrus Project Cirrus . Later, the Office of Naval Research lent support, the U.S. Air Force gave flight support, and the Weather Bureau provided consultants.

The primary discovery made by Project Cirrus was that seeding supercooled stratus clouds with dry ice or silver iodide cleared paths through the clouds, as the artificially nucleated snow swept all visible particles out of the cloud. Moreover, eddy diffusion from a narrow plane track spread the effect laterally for more than one mile.

Project Cirrus also tried to seed cumulus clouds, without much effect. In 1948, scientists conducted important cumulus-cloud-seeding experiments in New Mexico. On October 24 of that year, when a hurricane off the coast of Florida was seeded, Project Cirrus personnel reported that seeding had produced a pronounced modification of the cloud deck. Shortly after the seeding, the hurricane changed course and struck the coast of South Carolina and Georgia. Speculation that seeding the hurricane may have caused its course change caused hurricane-seeding experiments to be discontinued.

In August, 1947, the U.S. Weather Bureau decided to carry out its own experiments. The project, which spanned two years, was carried out in Ohio, California, and the Gulf states. The results of these investigations were in conflict with those of Langmuir and Schaefer.

When Luna B. Leopold Leopold, Luna B. and Maurice H. Halstead Halstead, Maurice H. seeded cumulus clouds in Hawaii, rainfall resulted. Langmuir surmised that the cloud had actually been seeded by water, which coated the dry-ice pellets. This led to Langmuir’s collision-coalescence theory of drop growth (a theory that had been postulated earlier by others), in which he speculated that warm clouds could be seeded with water to produce precipitation, and that giant hygroscopic nuclei could be used as well as water.

In 1948 and 1949, Langmuir studied tropical clouds in Honduras and investigated the work of Joe Silverthorne Silverthorne, Joe , a commercial cloud seeder working for the United Fruit Company. Silverthorne was interested in controlling rainfall and downbursts from thunderstorms that destroyed stands of fruit trees. On a flight, Langmuir dropped one pellet into a cloud and two into another, causing both to dissipate.

In 1950, Langmuir proposed two methods of seeding cumulus clouds. He postulated that self-propagating storms could be produced by dropping one pellet of dry ice in a cloud, and that by overseeding a cloud, the drops or ice crystals that formed would be too small to fall to the earth. Work with ground generators issuing silver-iodide smoke went on at the same time, but with mixed results. Eventually, a more successful technique was developed that involved seeding clouds with silver iodide from planes rather than from ground generators.


Following Langmuir’s and Schaefer’s research results, commercial operators continued to advance weather-modification techniques. Cloud seeding was done more cautiously from then on because of the legal ramifications regarding rights to water in the atmosphere.

Other long-term effects of the initial experiment included a knowledge of the downwind effect of cloud seeding—particularly inadvertent cloud seeding—on precipitation, and the awareness that cloud seeding, intentional or inadvertent, could reduce rainfall over cities (overseeding of cumulus clouds was known to have reduced rainfall over Florida). Furthermore, the experiment showed that it was possible to clear cold fog from airports. Weather modification
Cloud seeding

Further Reading

  • Battan, Louis J. Harvesting the Clouds; Advances in Weather Modification. Garden City, N.Y.: Doubleday, 1969. Discusses weather modification and cloud seeding for general readers.
  • Cotton, William R., and Roger A. Pielke. Human Impacts on Weather and Climate. New York: Cambridge University Press, 1995. Examines the effect of humans on changes in weather patterns. The first three chapters discuss cloud seeding in particular.
  • Fleming, James R. “Fixing the Weather and Climate: Military and Civilian Schemes for Cloud Seeding and Climate Engineering.” In the Technological Fix: How People Use Technology to Create and Solve Problems, edited by Lisa Rosner. New York: Routledge, 2004. Part of a collection that examines the use and abuse of technology, this article looks at “climate engineering,” or cloud seeding, as practiced by the civilian and military sectors.
  • Halacy, Daniel S., Jr. The Weather Changers. New York: Harper & Row, 1968. A popular look at weather modification and cloud seeding. Good account for the nonspecialist.
  • Hess, W. N., ed. Weather and Climate Modification. New York: John Wiley & Sons, 1974. An excellent history of weather and climate modification written for the nonspecialist. Some knowledge of math is helpful but not necessary. Presents differing opinions and the results of different studies.
  • Kwa, Chunglin. “The Rise and Fall of Weather Modification: Changes in American Attitudes Toward Technology, Nature, and Society.” In Changing the Atmosphere: Expert Knowledge and Environmental Governance, edited by Clark A. Miller and Paul N. Edwards. Cambridge, Mass.: MIT Press, 2001. Part of the Politics, Science, and the Environment series, this article examines public opinion on weather modification over time.
  • Mason, B. J. Clouds, Rain, and Rainmaking. New York: Cambridge University Press, 1975. A technical discussion of weather modification and cloud seeding. Some knowledge of meteorology is helpful.
  • “Project Cirrus: The Story of Cloud Seeding.” General Electric Review 55 (November, 1952): 8-26. An informative article outlining the major accomplishments of Project Cirrus.
  • Suits, C. Guy, and Harold E. Way, eds. The Collected Works of Irving Langmuir. Vol. 11. New York: Pergamon Press, 1962. The entire twelve-volume set gives a good assessment of Langmuir’s work. A knowledge of math is helpful. A previously unpublished article provides a detailed look into cloud-seeding experiments.

Soviets Adopt Stalin’s Plan for the Transformation of Nature

Acid Rain Changes Lake and Riverine Ecology

Greenhouse Effect Is First Predicted