THE USE OF CLOUDS IN FORECASTING 
and as recently discussed quantitatively by Byers and 
Battan [22, p. 173; 99]. 
The development of Cu congestus in the morning is 
to be used with caution in forecasting afternoon 
showers; because a strong inversion may stop them 
short of sufficient vertical development. Or, dry air 
aloft, usually revealed by fraying tops, may be tapped 
by the convection, thereby raising the cloud base and 
lowering the temperature and consequent vertical reach 
of the clouds. Similarly, as noted above, Acc frequently 
do not indicate a shower. 
Altocumulus Direction and Speed and Thunderstorm 
Movement. In a period when afternoon thunderstorms 
are occurring or are expected, the direction of motion 
and the apparent speed of the Ac should be noted. If, 
for example, the motion is from the west, any large 
cloud heaps or arched tops of thunderstorms in that 
direction should be watched closely. If the Ac have 
been moving fast, shelter should be sought soon, but 
if they have been going slowly, a distant storm may not 
arrive for two hours. Furthermore, if a growing thunder- 
storm is seen approaching apparently from the north- 
west, even if the Ac have been moving rapidly from the 
west, the storm is not likely to break as soon as might 
be expected, for the nearest portion of the storm is not 
coming toward the observer. Under such conditions, 
however, the squall will come from the west-northwest 
or northwest, and is likely to prevail for an appreciable 
interval before rain begins. 
Wind-Shift Line Marked by Clouds. When this con- 
dition is noted at either a squall line or a cold front, 
showers are likely. A wind-shift line at the earth’s 
surface is usually marked by some forced ascent of air; 
the appearance and approach of cloud forms due to 
forced ascent thus indicate coming change, often rain, 
within a few minutes to a few hours, depending on the 
distance to which the clouds can be seen and the rate 
of approach. If there is a well-defined cloud base, the 
height of which is estimated, successive measurements 
of the angular altitude of the base as it approaches will 
give a reasonably accurate estimate of the time of ar- 
rival of the wind shift. 
The approach of a strong wind-shift line may be de- 
tected an hour or more in advance by observing the thin 
white arch of the Cz border to the anvil top of the ap- 
proaching Cb. Prefrontal weather is so hazy that only 
this whitish arch will reveal the presence of a moderately 
distant Cb, for the shadowed air under the heavy anvil 
will be invisible behind the sunlit hazy blue air near the 
observer. Thus this part of the sky will appear to be 
clear and will resemble the blue sky above the C7 arch. 
When a cold front has passed, the lifting effect of the 
invading cold air mass continues to make clouds form 
in middle and upper levels. Though the bank of clouds 
marking the rainy area associated with the front con- 
tinues to recede in the east or southeast, new clouds, in 
bands parallel to the front, continue to form. These 
new bands will usually not form as rapidly as the gen- 
eral movement of the front carries the whole cloud 
system away. Indeed, in the daytime, solar heating may 
throw the western or northwestern edge more rapidly 
1173 
eastward or southeastward than the rate of progress of 
the front. In the late afternoon and evening, however, 
the cooling of the moist layers by radiation may, in 
the case of slowly moving fronts particularly, make the 
cloud edge retrograde, giving an apparent threat of 
renewed precipitation which, of course, is quite pos- 
sible if the front is not far away and if it has some waves 
in it. 
A daytime phenomenon of the clearing-off zone, which 
Dr. Aili Nurminen has told the author not infrequently 
makes airport forecasts of improving ceilings fail, is 
the formation of an abundance of very low fractocumulus 
as soon as the middle and upper clouds permit moderate 
insolation to reach the wet ground. Fortunately, these 
soon break, for they so reduce the insolation that con- 
vection is weakened. Though more clouds will then 
form owing to renewed insolation, their bases will be 
higher. 
Difficulties rn Forecasting Fronts in Tropics. In the 
tropics, where fronts are usually slow-moving and often 
stationary for long periods, the forecasting of frontal 
passages, which are the chief elements of change in those 
latitudes, is difficult, as Deppermann [34] very reluc- 
tantly admits. Fronts at Manila come mainly from the 
north, and the upper-air changes due to the overriding 
of the northers by the trade will occur north of the front 
and not near Manila. The rare returns of fronts from 
the south may be heralded by changes in the trade 
wind aloft. The equatorial front has little overrunning, 
for the temperature differences between the trades and 
the southwest monsoon are slight. Even the pre-typhoon 
sequence of clouds is not perfectly regular, and almost 
all the cloud forms found in typhoons can also appear 
on days of strong northers or of many thunderstorms. 
Synoptic Types of Sky: How Synoptic Cloud Data May 
Aid Weather Forecasting 
We in the United States pay less attention to states 
of the sky and cloud sequences than do meteorologists 
of western Europe, because, except on the Pacific Coast, 
we have a fairly adequate network of stations, and so 
do not have to depend upon cloud forms and relation- 
ships to give us an idea of weather conditions to the 
west. Nevertheless, if we do not use the cloud portion 
of the weather picture, we are depriving ourselves of 
this considerable advantage in analyzing the weather 
situation and in following current trends. Schereschew- 
sky [84] emphasized this strongly, as did Bigelow [7] 
and Ley [60] before him. 
The forecaster’s situation is still happier, however, 
when the usual synoptic surface and upper-air data 
are supplemented with accurate and frequent cloud 
observations. Of the nineteen elements plotted on the 
U.S. Daily Weather Maps, seven relate to cloud con- 
ditions, and three of these are devoted to thirty separate 
indications of clouds and their trends at various levels. 
The thirty indications are divided into three groups 
of ten, Cu, Cm, and Cr, each roughly representative of 
conditions at high, middle, and low levels. The sequence 
of numbers in each group is generally that from better 
to worse weather as a storm approaches, ending with the 
