558 
slopes it is subjected to vertical stretching (horizontal 
convergence). A similar phenomenon is observed in 
August but is interrupted at lower latitudes by the 
upper-level anticyclone over the United States. Similar 
charts, prepared for other levels, might throw consider- 
able light on many problems of the dynamics of the 
general circulation. 
If the normal fields of divergence at 10,000 ft are 
fairly representative of the average divergence in lower 
atmospheric levels where there are no mountains, it is 
possible to obtain some idea of the normal fields of 
vertical motion. Thus there is probably a region of 
gradual subsidence over the eastern United States and 
a region of gradual ascent over most of the Atlantic. 
Similar regions are indicated over the Pacific. The charts 
also indicate subsidence in eastern portions of the sub- 
tropical highs and ascent in western portions, conclu- 
sions which are in good agreement with the Bjerknes 
subtropical models [5] and with observed precipitation 
regimes. 
It is recognized that in order to obtain a complete 
explanation of the normal state of the general circula- 
tion it is necessary to locate and determine the magni- 
tude of the sources of heat and moisture in the entire 
world’s atmosphere, for it is these sources which supply 
the necessary energy to drive the atmospheric heat 
engine. Some of the most recent work on this topic has 
been done by Wexler [54] and by Jacobs [25, 26]. 
A & 
590 
THE GENERAL CIRCULATION 
study, Winston and Aubert,’ estimating the effects of 
vertical motion, find that Wexler’s method appears to 
give good estimates of total heating in regions where this 
heating or cooling is large. Thus, while part of the heat 
gain shown over eastern North America in Fig. 9 may be 
attributed to adiabatic warming through subsidence, 
it is not possible to explain in this manner the centers 
of strong heating appearing off the coasts of the con- 
tinents. Apparently these represent heat sources in- 
strumental in driving the circulation of the Northern 
Hemisphere. Similar reasoning applied to other areas 
suggests that the regions of cooling over the north- 
eastern parts of oceans are heat sinks, while subsidence 
accounts for the apparent heat sources over the south- 
eastern Atlantic and Pacific. 
The studies discussed above do not consider individ- 
ual heating processes affecting the circulation, for 
example, radiation, condensation, sensible heat ex- 
change with the earth’s surface, and mixing. Contri- 
butions to these aspects of the problem have been made 
by Jacobs [25, 26] who calculated for the air overlying 
oceanic regions the normal seasonal heating due to 
condensation and to sensible heat exchange with the 
surface. Jacobs found that the air was being heated 
from below in both the western and the northeastern 
portions of the oceans. The amounts of heating he found 
are in general agréement with the total heating as given 
by Wexler only in the western portions, but im the 
northeastern portions Wexler’s values (showing net 
RATE OF HEAT GAIN 
ai CAL cM=2 pay"! 
“ie ——-—-—RATE OF HEAT LOSS 
+ _IN GAL cm72 Day7! 
Fie. 9—Regions of normal heat gain and loss for the layer from sea level to 10,000 ft for February. (After Wexler [54].) 
Using data for the Northern Hemisphere, Wexler 
estimated the normal regions of heating and cooling of 
the air in the layer between sea level and 10,000 ft for 
the months of February (Fig. 9) and August. As Wex- 
ler points out, more exact values of the total heat gain 
or loss would be possible if the effects of vertical motion 
were considered. But the quantitative evaluation of this 
term is difficult. In an attempt to extend Wexler’s 
cooling) must be accounted for in some other manner, 
perhaps by radiation or mixing. Indeed the importance 
of horizontal mixing in cooling northward-moving air 
4. Part of the work of this continuing project at the Ex- 
tended Forecast Section of the U. S. Weather Bureau will be 
published shortly. A further study is being made of normal 
and anomalous sources and sinks for other months and other 
levels. 
