THE AEROLOGY OF EXTRATROPICAL DISTURBANCES 
By E. PALMEN 
Academy of Finland 
The present article deals mainly with the role of 
extratropical disturbances as links in the general at- 
mospheric circulation and as cells for meridional ex- 
change of air masses. Because of limited space we can- 
not discuss the subject in detail. Since there already 
exists a large amount of literature on the structure and 
behavior of cyclones im the surface layers, most em- 
phasis has been placed in this article on the structure 
of disturbances in the free atmosphere. Most of the 
ideas and viewpoints expressed here can be found in 
the extensive writings dealing with extratropical cy- 
clones. However, it has not been possible to refer 
directly to more than a portion of this material. 
This article was prepared in final form while the 
writer was under appomtment as visiting professor at 
the University of Chicago. The writer is indebted to 
Mr. C. W. Newton for his valuable help in preparing 
many of the figures in the article and especially for his 
analysis of the cyclone of November 17-19, 1948. 
CHARACTERISTIC TEMPERATURE AND WIND 
DISTRIBUTION IN THE WEST-WIND 
ZONE 
The formation of extratropical cyclones is connected 
with fronts separating air masses of different origin.! 
The most important front is that which separates air 
masses of polar origin from those of tropical origin. 
This “polar front” can, at least during the colder season, 
be followed relatively easily around half or more of the 
Northern Hemisphere on daily synoptic surface maps. 
The breaks im the polar front are of great importance. 
In their classical paper on the polar-front theory, J. 
Bjerknes and H. Solberg [7] pointed out that these 
breaks in the polar front are necessary in order to per- 
mit an exchange of air masses between the polar and 
tropical source regions. 
From studies of synoptic charts for different levels 
one can conclude that the outflow of polar air from the 
polar source region occurs particularly in the regions 
between two polar fronts separated by a cold anti- 
cyclone. The polar outflow here appears in the southern 
parts as a relatively shallow layer of subsiding air 
masses. On the other hand, the inflow of tropical air 
into the polar region at higher latitudes usually appears 
as currents of warm air in the upper atmosphere. This 
can be seen from statistical studies of the frequency of 
the two principal air masses at different levels [35]. 
Frontal analyses of upper-air charts indicate that the 
area occupied by tropical air increases with height at 
all latitudes. Since the characteristic difference between 
1. If taken in the broadest meaning of the word, cyclones 
or depressions of different types can form without any pre- 
existing fronts or frontal zones. These types of irregular, 
generally weak depressions will not be discussed here. 
polar and tropical air at a given latitude could not be 
maintained for a long time without advection, the only 
possible explanation for the above-mentioned fact seems 
to be the existence of a systematic meridional circula- 
tion of the type outlined here. 
It is necessary to emphasize that the meridional cir- 
culation appears as a statistical fact. The average 
meridional circulation is naturally not the same at all 
longitudes nor at all times during a certain season. The 
circulation has a tendency to break up into several cells 
with quasi-vertical axes, depending upon the geographi- 
cal distribution of continents and oceans and the forma- 
tion of disturbances [11]. Thus the paths of individual 
air parcels are extremely complicated and the different 
air parcels carry out various kinds of complicated os- 
cillatmg movements before they have performed a com- 
plete meridional circulation. 
If there is a statistically opposite meridional move- 
ment in the lower and upper troposphere, obviously 
there must also be a surface of zero average meridional 
movement, separating the northward flow in the upper 
troposphere from the southward flow in the lower 
troposphere. This surface has a tendency to become a 
frontogenetical surface. Since the meridional move- 
ments discussed here are average displacements for 
longer periods, the separating surface naturally cannot 
have the character of a real front. However, synoptic 
experience shows that in nature there is a strong tend- 
ency to concentrate the contrasts between the air masses 
of tropical origin and those of polar origin in a rela- 
tively thin layer of transition. This transitional layer 
must have a certain inclination toward the horizontal 
surfaces because of the earth’s rotation. On a nonrotat- 
ing earth, frontal surfaces of the type observed in the 
atmosphere, with an inclination of the order of 
149-1400, could never form. 
Tt is obvious that there must be a transformation of 
polar air into tropical air at low latitudes and that the 
opposite must be true at high latitudes. The outflow 
of polar air into the subtropics takes place in the lower 
troposphere and the inflow of tropical air into the 
polar regions takes place in the middle and upper 
troposphere, and perhaps partly in the lower stratos- 
phere. Therefore there cannot be any continuous frontal 
surface around the whole hemisphere in the vicinity 
of the earth’s surface or in the upper atmosphere. 
However, in other parts of the troposphere there is 
nothing in principle against the idea of having a con- 
tinuous polar front around the entire hemisphere.” 
2. The concept of a front is here used in the broad sense; it is 
obvious that a distinct upper polar front can be expected only 
in special situations favoring an extreme concentration of 
the air-mass distinctions. 
599 
