RELATION OF ARTIFICIAL CLOUD-MODIFICATION TO THE PRODUCTION 
OF PRECIPITATION 
By RICHARD D. COONS and ROSS GUNN 
Physical Research Division, U.S. Weather Bureau 
Introduction 
With the important discovery in 1946 that super- 
cooled cloud elements could be artificially converted to 
ice crystals by introducing pellets of carbon dioxide 
snow [11, 12], the possibility of human control of 
weather seemed imminent. Claims and speculation con- 
cerning the degree of possible weather control and the 
probable amount of artificially induced precipitation 
were rife, and meteorological opinion ranged from the 
one extreme “‘of academic value only” to the other “of 
great economic and military importance.”’ The more 
favorable claims were supported in part by a few in- 
completely documented single experiments conducted 
in 1946 and 1947. These claims excited the interest of 
the public and stimulated demands for immediate 
weather control, drought relief, and storm diversion 
and dissipation, even though at the time there was no 
definite objective evidence that these were possible. 
Alfred Wegener [15] was the first to suggest that the 
coexistence of ice and water in supercooled clouds led to 
colloidal instability resultmg im rapid growth of ice 
particles. The possibility of modifying and producing 
rain from supercooled clouds by adding sublimation 
nuclei was foreseen by Bergeron [1, 2] and Findeisen 
[8]. Until 1946, however, no method was known of pro- 
ducing the necessary ice-crystal nuclei artificially. Har- 
lier experiments by Veraart [13] m Holland in 1930 
in which he dropped solid carbon dioxide, among other 
things, mto supercooled clouds must have produced 
such nuclei, although Veraart at that time did not 
recognize this possibility. He was credited with produc- 
ing slight amounts of rain on several occasions. How- 
ever, because of his sweeping claims, even his positive 
results were discredited. Not until Schaefer and Lang- 
muir [12] demonstrated (1) that a single solid carbon 
dioxide pellet could produce prodigious numbers (of 
the order of 10'*) of sublimation nuclei in its fall through 
a supercooled cloud, and (2) that, in actual field tests, 
a number of CQ pellets dropped into a supercooled 
cloud converted it largely into ice crystals, was it pos- 
sible to investigate the importance of artificial nuclea- 
tion in producing rain and modifying clouds as pre- 
viously suggested by Bergeron and Findeisen. More 
stable sources of sublimation nuclei have also been 
found suitable. Silver iodide in particular has been used 
most successfully [14]. It should be noted that Heverly 
[9] has demonstrated in the laboratory that spontaneous 
freezing of water droplets in supercooled clouds may 
be a more important natural precipitation instrument 
than the sublimation of water vapor onto suitable active 
nuclei. According to Heverly’s results, the effect of dry 
ice in producing colloidal instability could possibly be 
due either to the resulting spontaneous freezing of the 
supercooled droplets or to the production of myriads 
of ice-crystal germs from the water vapor in the cloud. 
In any case, whatever the effect of the dry ice, there is 
no question that it is capable of almost completely 
converting seeded portions of supercooled clouds into 
ice particles. The optical effects of such conversions 
are startling. In Fig. 1 the reflection of the sunlight 
from the ice-crystal cloud produced by seeding is seen 
to be brilliant when compared to the light reflected from 
the adjacent supercooled water-droplet clouds. 
Fre. 1—Reflection of sunlight on aluminum aircraft wing, 
ice-crystal cloud, and supercooled water-droplet cloud. Note 
that the reflection from the ice-crystal cloud is nearly as bright 
as that from the wing. 
Bergeron [3] has fully discussed the suitability of 
the various types of rain-producing clouds for artificial 
release of precipitation. Briefly, the following conditions 
must be met in a cloud or cloud system in order for arti- 
ficial nucleation to be effective in the modification of 
clouds and the production of precipitation: (1) The 
cloud must contain water droplets at a temperature be- 
low freezing, while there are still no ice crystals present 
in the immediate vicinity or falling through the cloud : 
(2) it must have some minimum thickness (arbitrarily 
about 1000 m) in order for the amount of moisture that 
could be converted into precipitation to be appreciable; 
(3) there probably should be some vertical instability 
in order to allow the growth, if any, of the cloud to 
spread the ice crystals as a result of the heat liberated 
by the freezing of the water (this would be accentuated 
in the presence of large-scale horizontal convergence 
that would provide a continuing source of moisture 
into the seeded area). Less important factors deter- 
mining the possibility that precipitation will reach the 
235 
