72 AIR MASS ANALYSIS 
The Ice-Nuclei Theory of Rainfall 
When a cloud top grows a consider- 
able distance into region above the 
0°C level, more and more ice crystals 
will form, creating a layer of water 
and ice particles and perhaps still 
higher one with ice crystals only. 
Bergeron developed a theory (later 
elaborated by Findeisen) that large- 
dropped rain will not be precipitated in 
considerable amounts until the top of 
the cloud is partly composed of ice 
crystals, because the difference in 
vapor pressure over ice and water 
leads to rapid growth of the ice 
crystals by evaporation from the 
drops, the ice crystals then falling out, 
melting and taking on more water on 
the way down as rain. Other clouds are 
more colloidally stable and give only 
fog or drizzle. This theory seems to 
agree with observations in middle and 
high latitudes, but is not yet accepted 
as universally true, as there is much 
evidence from the tropics and sub- 
tropics that certainly it is not always 
applicable there. 
If and where the principle holds, 
however, then the most favorable con- 
ditions for the development of thun- 
derstorms would exist when the dis- 
tance between the lowest level where 
convective energy would be freed and 
the ice-crystal level is great and the 
amount of latent energy is large. Un- 
der such conditions the rising particles 
of air would have sufficient energy to 
reach well above the ice-crystal level, 
and at the same time the lower air 
particles would be accelerated up- 
ward for a long time, leading to high 
velocities and release a maximum of 
heat from condensation.—R. G. S. 
Hail Forecasting 
A recent study made by United Air 
Lines meteorologists indicates that in 
the central and eastern United States, 
large hail falls mostly in showers oc- 
curring between noon and 9:00 P.M. 
and especially between 3:00 and 6:00 
P.M. Over half the hail falls were 
due to frontal activity, but insola- 
tional heating during the day ap- 
parently intensified the frontal “‘trig- 
ger action” to a great extent. The 
adiabatic charts and Rossby diagrams 
for hail days were studied but did not 
give very definite criteria for fore- 
casting these hail storms. This was 
most likely due to the fact that the 
airplane soundings analyzed did not 
reach over 15,000 or 17,000 feet, and 
at those elevations the positive energy 
areas were usually either at their 
maximum width or still wide open, so 
that the real extent of the energy 
areas could not be taken into account. 
With the radiosondes now used this 
deficiency is eliminated. The follow- 
ing empirical results from the study, 
however, have a certain forecasting 
value :— 
No falls of large-hail were reported 
in thunderstorms which occurred 
when the base of the conditional or 
convective instability region was 
above 7,000 feet. (This region began 
anywhere between 3,000 and 14,000 
feet for hailless thunderstorms, but 
between 3,000 and 7,000 feet for 
storms with large hail.) 
The proportion of thunderstorms 
producing large hail was one per 400 
when the lapse rate (°F per 1,000 ft) 
was greater than 4.5°F in the con- 
vective or conditional instability re- 
gion as indicated by the nearest 
sounding. The frequency was only 
one-third this great (one storm in 
1,200) when this lapse rate was 
smaller than 4.5°F and the base of 
the instability level still below 7,000 
ft. The lapse rate referred to was 
generally between 3.5° and 4.5°F for 
non-hail storms, but anything between 
a and 5.5° for storms with large 
ail. 
