Cloud seeding

E. Linacre and B. Geerts


Cloud seeding in Australia

Cloud-seeding experiments were carried put in Australia from 1947 (Note 9.E) till 1994 (1). In the agricultural areas of southwestern and eastern Australia, there has long been an interest in rainfall enhancement by means of cloud seeding. Cloud seeding is most promising in clouds with cloud tops warmer than -15° C to -20° C, because these clouds may be largely or entirely supercooled, in which case the introduction of ice nuclei would produce snow flakes, through the Bergeron-Findeisen process. That is, ice crystals would grow rapidly at the expense of supercooled droplets because the vapour pressure over ice is lower than that over water.

Commercial seeding has been deemed unfeasible over Australia’s arid inland plains, because most rain there falls from clouds already containing ice. The same applies to rain falling during the summer in northern Australia, where variability of rainfall makes it harder to demonstrate statistically significant gains in precipitation from cloud seeding. Studies in western Victoria (in southeast Australia) shows that seeding rarely is practicable with frontal clouds or when the winds are from the southwest, because the cloud water content is small and/or the cloud tops are too warm. Unfortunately that leaves only a few days each year suitable for seeding.

On the other hand, orographically lifted stratiform clouds are more likely to contain large concentrations of supercooled liquid water. Therefore cloud seeding may be commercially viable in mountainous areas. For instance, two experiments in Tasmania showed useful rainfall enhancement when cloud-top temperatures are between -10° C to -12° C. The current position is that Tasmania is the only state in Australia where the authorities are convinced of the economic value of cloud seeding. An estimate in 1994 gave the cost of a cloud-seeding operation in Tasmania as $645,000, providing an additional 55 mm of rain over 6 months. This represents a benefit-cost ratio of 13:1.

Experiments in various areas of mainland Australia all showed an unexpected decline over 3-6 years in the effectiveness of cloud seeding an area. The reason is not clear. Keith Bigg proposed that it arises because silver on vegetation, from silver iodide used in previous seedings, which multiplies the formation of bacteria which themselves act as nuclei in rain formation, confusing comparisons with rainfalls from unseeded, control areas (2). In other words, rainfall gradually increases in the control areas because of persistent contamination from the seeded areas, reducing the superiority of the seeded areas. However, there is no explanation of how the bacteria could reach cloud level. Also, there has been no apparent increase in rainfall in the ‘control’ areas.


Commercial cloud seeding elsewhere

Cloud seeding is a profitable commercial operation in various places in the USA, mainly in the Sierra Nevada, the Colorado Rockies, and the Great Plains. Experiments in the Colorado Rockies in 1960-70 showed that rainfall is increased when the 700 hPa winds are from the southwest and 500 hPa temperatures higher than -20° C. A 1995 field experiment in Arizona (3) highlighted the need to understand the detailed airflow structure over topography to successfully seed clouds. In this field experiment, a high amplitude gravity wave was shown to occur downwind of a mountain ridge during winter storms, and the seeding of the ascending branch of that wave was more effective than the seeding of higher terrain further downstream. Two experiments in Israel in 1961-75 indicated about 14% increase, but a third showed a decrease, unless one excluded days with desert dust in the air.


Diffusing tropical cyclones

Fig 1: Illustration of the hypothetical effect of the seeding of a tropical cumulus cloud with silver iodide. On the left, silver iodide is released into a partly supercooled cloud. On the right, this cloud is invigorated into a cumulonimbus by the latent heat released as the boundary between liquid and frozen hydrometeors (blue horizontal line in the cloud on the left) moves down to the 0° C isotherm (green horizontal line). Net freezing releases latent heat.

In the 1960s the US National Weather Service staged a project STORMFURY which aimed to weaken hurricane intensity by introducing ice nuclei at low levels in the updraft region around the cyclone’s eye, i.e. in the eyewall. The theory was that the ice nuclei would trigger freezing down to the 0° C isotherm, rather than at some height above the freezing level (see Fig 1) (4). Therefore the latent heat released by freezing would be concentrated at a lower level, and this energy, together with the latent heat of condensation (which also is released mainly at low levels), would add buoyancy, strengthen the updrafts, and ensure more low-level convergence. This short-lived, but rapid intensification of the hurricane would widen the diameter of the eyewall, which in turn would reduce the wind speed (i.e., the conservation of angular momentum). The Project STORMFURY could not prove the validity of this hypothesis, and hurricanes are no longer seeded (5).


A great review of cloud seeding can be found at



  1. Ryan, B.F. and W.D. King 1997. A critical review of the Australian experience in cloud seeding. Bull. Amer. Meteor. Soc. 78, 239-53.
  2. Bigg, E.K. and E. Turton 1988. Persistent effects of cloud seeding with silver iodide. J. Appl. Meteor. 27, 453-7.
  3. Klimowski, B.A., R. Becker, E.A. Betterton, R. Bruintjes, T.L. Clark, W.D. Hall, B.W. Orr, R.A. Kropfli, P. Piironen, R. Reinking, D. Sundie, and T. Uttal, 1998. The 1995 Arizona Program: toward a better understanding of winter storm precipitation development in mountainous terrain. Bull. Amer. Meteor. Soc., 79, 799-813.
  4. Malkus, J. S., and R. H. Simpson, 1964: Modification experiments on tropical cumulus clouds. Science, 145, 541--548.
  5. Willoughby, H. E., D. P. Jorgensen, R. A. Black, and S. L. Rosenthal, 1985: Project STORMFURY, A Scientific Chronicle, 1962-1983, Bull. Amer. Meteor. Soc., 66, 505-514.