1202 
The most severe icing conditions occurring anywhere 
are found in the upper half of tall cumulus congestus 
clouds just before they reach the cumulonimbus stage. 
Ordinarily, such clouds comprise only a small portion 
of the total air mass and therefore they can usually be 
circumnavigated. Also, the limited horizontal extent of 
cumulus towers limits the amount of ice that can be 
collected during a single passage through a cloud. Be- 
cause of the removal of liquid water by precipitation, 
icing conditions in cumulonimbus clouds are usually 
much less severe than in cumulus congestus clouds; 
however, conditions in cumulonimbus are extremely 
variable, and sometimes very heavy icing may be en- 
countered for short distances. 
Stratus and Stratocumulus Clouds. Stratus and strato- 
cumulus clouds ordinarily contain moderate to low 
concentrations of liquid water, although fairly high 
values sometimes occur near the tops of unusually 
thick layers. These clouds are rarely characterized by 
large values of drop diameter. Although the icing con- 
ditions in stratus and stratocumulus clouds are usually 
light to moderate in intensity, they present one of the 
chief icing hazards to aircraft operation, since the cloud 
layers frequently cover large areas and it is not pos- 
sible to ascend or descend without flying through them. 
Also, traffic control procedures may occasionally re- 
quire prolonged flight in a stratocumulus layer. The 
intensity of icing usually increases from near zero at the 
base of the cloud layer to a maximum just below the 
top, the maximum intensity being proportional to the 
thickness of the layer. The factor which limits the 
intensity of icing is the limited thickness of the cloud 
layers. Individual layers of stratus or stratocumulus 
clouds seldom exceed 3000 ft in vertical extent. For 
layers of a given thickness, the liquid-water content is 
greater in warmer clouds, hence the greatest rates of 
ice accretion will be found at temperatures only a few 
degrees below freezing. This relation does not apply to 
individual clouds, however, since in any particular 
cloud layer the icing is most severe in the upper, 
colder portions. 
When light or very light snow falls from a stratus or 
stratocumulus layer it is an indication of decreasing 
liquid-water concentration and icing imtensity; when 
moderate or heavy snow occurs, the liquid water dis- 
appears very quickly and little or no icing is to be ex- 
pected. 
Altocumulus Clouds. Icing conditions in altocumulus 
clouds are generally similar to, but milder than, those 
in stratocumulus clouds, as the temperatures are usually 
lower and the average cloud thickness is less. Large 
values of drop diameter occur more frequently in alto- 
cumulus clouds with the result that icing conditions 
are occasionally considerably more severe than would 
be expected from the thickness of the layer. Ordinarily, 
however, only light and occasionally moderate condi- 
tions are encountered in altocumulus clouds and these 
are usually not difficult to avoid by a change of alti- 
tude. 
Altostratus Clouds and Altocumulus Associated with 
Altostratus. Altostratus clouds usually consist of thick 
CLOUDS, FOG, AND AIRCRAFT ICING 
and extensive cloud masses which are composed almost 
entirely of ice crystals and therefore do not cause icing. 
This fact is of the greatest importance in forecasting 
icing conditions. The failure of forecasters to realize 
that altostratus clouds and the large precipitation areas 
usually composed predominantly of altostratus con- 
tain little or no supercooled liquid water is probably 
responsible for more errors in forecasting icing condi- 
tions than any other factor. 
Around the edges of altostratus cloud systems, es- 
pecially on the side where air is flowing upward into the 
system, there is frequently found a rather complex 
grouping of altocumulus layers. These usually merge 
with the altostratus, losmg their liquid water as they 
become impregnated with ice crystals. Frequently, 
rather large cloud areas of mixed composition are 
formed in this way, for though a mixture of ice crystals 
and cloud drops is an unstable condition and therefore 
transient, a continuous supply of freshly formed alto- 
cumulus clouds is often available. Usually only light 
icing occurs in mixed clouds of this type. 
Flight Planning and Dispatching in Relation to Icing 
Conditions. The consideration that should be given to 
icing in flight planning is largely a function of the ice- 
protection equipment on the airplane. Aircraft without 
deicing equipment of any kind should avoid all sub- 
freezing clouds if possible. Aircraft equipped with rub- 
ber-boot deicers should avoid prolonged flight in con- 
tinuous stratus or stratocumulus layers by a suitable 
change in altitude and should circumnavigate heavy 
cumulus clouds whenever practicable. Aircraft having 
thermal ice-prevention systems can safely fly in a great 
majority of icing conditions, but if exceptionally severe 
or prolonged icing conditions are encountered, some 
changes in altitude or flight path may be advisable. 
It should not be necessary to postpone or cancel flights 
because of reported or expected icing conditions when 
adequate thermal ice protection is available. 
Because the occurrence of icing conditions is usually 
quite variable both in space and time, the practice of 
prohibiting flights in certain areas as a result of re- 
ported icing conditions is of doubtful validity. Such 
restrictions should not be applied to aireraft with ade- 
quate thermal ice-prevention systems. 
Problems Requiring Further Research 
One of the outstanding problems in the study of icing 
conditions is that of obtaining satisfactory measure- 
ments of the meteorological quantities involved. Much 
effort has been applied to this problem and noteworthy 
results have been achieved, but much still remains to 
be done. For example, there is a need for a more con- 
venient and dependable means of determining the dis- 
tribution of drop sizes in clouds, and also for a more 
accurate means of measuring mean-effective and maxi- 
mum drop diameters when these exceed 30 to 40 py. 
There is also a need for simple and dependable auto- 
matic instruments for recording liquid-water concen- 
tration and average drop size, instruments suitable 
for use during routine flights to obtain statistical data. 
One aspect of the icing problem in which our present 
