AEROLOGY OF EXTRATROPICAL DISTURBANCES 
polar and tropical air higher up in the troposphere. 
Many times the front is especially well marked only on 
the southwestern side of the large upper trough. The 
slope of the front over the western parts of an upper 
trough is smaller than over the eastern parts. Especially 
weak is the slope in the region above southwestern 
parts of the cold surface high, where the cold air ap- 
pears as a shallow mass. A typical case has been studied 
in detail by Palmén and Nagler [44]. 
It might be pomted out that m the region of the 
surface anticyclone there is often a tendency toward the 
formation of a new surface front between the colder 
polar air (continental air or “arctic air”) in the north- 
east and the somewhat warmer polar air (maritime air) 
in the southwest. This phenomenon is especially com- 
mon over the United States, where the mountains 
prevent the cold surface air from penetrating to the 
western coastal regions. Thus the continental anti- 
cyclone is ordinarily not as symmetrical as it has been 
drawn in Fig. 5; on the contrary, the anticyclone is very 
often split mto two parts separated by a secondary 
front, making possible the formation of disturbances 
similar in character, but weaker than those on the 
polar front. This secondary front has been disregarded 
in our schematic figures in order to avoid complications.® 
The general flow pattern at the surface and at 500 mb 
presented in Figs. 5 and 6 cannot be fully understood 
if the vertical components of the air motion are not 
introduced into the picture. In Fig. 6 the upper warm 
and cold air masses have an almost parallel geostrophic 
flow. Upon this geostrophic flow is superposed a com- 
plicated pattern of nongeostrophic motions. The pattern 
of vertical movement in upper troughs and associated 
cyclones has been subjected to a careful study at New 
York University by Miller [86, 37] and Fleagle [24]. 
According to their results, there is, on the average, de- 
scending motion on the west side of the upper trough 
and ascending motion on the east side. This vertical 
motion is accompanied by low-level divergence and 
upper-level convergence on the west side, and low-level 
convergence and upper-level divergence on the east 
side. On the average, the vertical components reach 
their maximum values at the level of nondivergence 
(which is not necessarily horizontal). In the divergence 
field the vorticity of the air parcels moving from west 
to east undergoes systematic changes; the absolute 
vorticity has its highest value at the trough line and 
its lowest value at the ridge line. 
This relatively simple pattern for vertical motion 
and divergence, however, does not describe very well 
the real movement in a disturbance which necessarily 
must contain a component of an zrreversible process, as 
has been pointed out earlier. Studies of synoptic charts 
8. This secondary front is usually called the ‘“‘arctic front”’ 
in the United States because the cold Canadian continent is 
regarded as belonging to the arctic source region in winter. In 
Kurope a similar front often separates the maritime polar air 
masses from the continental polar air masses over the eastern 
parts of the continent. Since the source region of the latter air 
masses in Hurope is in the east, not in the north, the corre- 
sponding front is not identified with the arctic front. 
609 
indicate that there is on the average an outflow of cold 
air from the surface anticyclone and consequently also 
a general subsidence in the cold air, extending very far 
up into the upper cold trough. The cold surface anti- 
cyclone and the cold part of the upper trough represent 
regions where cold polar air flows to lower latitudes. 
In order to maintain contimuity the same amount of 
warmer air must flow into the polar regions. The out- 
flow of cold air on the average must represent a de- 
scending current; the inflow of warm air, an ascending 
current. A vertical circulation therefore is superposed 
upon the general picture of the horizontal air movement. 
This circulation is energy producing and serves to 
maintain the kinetic energy of the westerlies. 
In order to obtain a general picture of the three- 
dimensional movement im synoptic situations of the 
type discussed here, it would be necessary to follow the 
three-dimensional air trajectories for air parcels in 
different parts of our picture. That could to some extent 
be done by methods developed in the above-mentioned 
investigation at New York University. However, in 
order to give results applicable to a general cyclone 
theory, the method should be combimed with a careful 
three-dimensional frontal analysis and applied to synop- 
tic situations of well-defined types. 
Since no such investigations have yet been made, we 
must restrict the discussion to some rather general 
qualitative results concerning the three-dimensional tra- _ 
jectories of air parcels in selected parts of our sche- 
matic Figs. 5 and 6. 
An air parcel initially situated at pomt A in Fig. 6 
is a polar air parcel. It moves with a descending com- 
ponent relative to a warm air parcel at pomt C of the 
same 500-mb chart. Both parcels are descending as 
long as they are on the west side of the trough line. 
The cold air moves along a path of the type marked in 
Fig. 6, where the successive positions and pressures are 
indicated by small circles. The trajectory starts out with 
a slight cyclonic curvature, corresponding to the cy- 
clonic curvature observed in the upper trough, but the 
curvature changes gradually to an anticyclonic one 
when the air parcel reaches levels where dw/dz < 0. 
At the end of the trajectory the polar air moves anti- 
cyclonically as a shallow layer of air which rapidly 
loses the properties of a cold air mass because of heating 
from below and adiabatic heating due to subsidence. 
The warm air parcel at pomt C at first descends on 
the west side of the trough, but gradually overtakes the 
trough line and then starts ascending. At the time when 
the cold air flows out in the subtropics as an anticy- 
clonic surface current the warm air parcel has moved 
very far to the northeast and has eventually left the 
upper trough in which it started. There is no possibility 
of giving the exact position of the air, but it follows the 
general geostrophie flow in the upper troposphere with 
an average tendency to deviate a little to the left, 
because of the average southward displacement of the 
polar air underneath. 
If we assume that after thirty-six hours the cold air 
parcel is at the pot marked by 800 mb in Fig. 6, the 
mean velocity since it started from point A is about 
