556 
in February and March, the zonal index reaching a 
minimum around the beginning of March. Clearly such 
a minimum could not show up in monthly mean values 
such as are represented in Fig. 7. But averages com- 
i NORMAL INDICES 
M SEC! (COMPLETE NORTHERN HEMISPHERE) 
4 
2 
eval + 
4} POLAR EASTERLIES 
WW 
> 6 
=) | 
46 
a ZONAL WESTERLIES [laa 
4 
2 
©| SUB TROPIGAL EASTERLIES 
6 
4 
2 
o Ol POLAR 
za 12 Doo i 
= 10 
- 8 
an 
Sa 
oa) 
= 6 
6 TEMPERATE 
fo) 
ft 10 [ 
8 
6 
4 
£ | SUB TROPICAL in 
J F M A M J J A 5S © N D J 
Fie. 7.—Normal Northern Hemisphere monthly indices. 
puted from five-day mean charts worked up for ten- 
day intervals during the six years 1944-49 (when reliable 
upper-air data were available) strongly suggest the 
reality of this “global singularity.”” Another period of 
similar data (1932-39) worked up in part with the help 
of extrapolation techniques [59] also suggests the real- 
ity of this sharp dip in the normal zonal index. More 
concerning this topic will be said in a following section. 
As pointed out earlier, the state of our knowledge of 
the observed normal state of the general circulation is 
still quite inadequate. Our deficiencies lie particularly 
in the absence of a long period of record of upper-air 
data over large areas of the Northern Hemisphere and 
over most of the Southern Hemisphere. Not much can 
be done for many years to remedy the “long-period” 
part of this deficiency. But it would appear well worth 
while to obtain, if only for one year, a reasonably com- 
plete series of daily soundings well into the stratosphere 
covering the entire world. From this material many of 
our blind spots would be cleared up. It might even be 
possible to lmk up interactions between the circulations 
of the hemispheres. Also, secrets of behavior of atmos- 
pheric circulation might conceivably be easier to find 
in the Southern Hemisphere where topographically pro- 
duced distortions are comparatively lacking. In terms 
of modern expenditures for scientific work the cost of 
THE GENERAL CIRCULATION 
such a truly international observation year would not 
appear to be prohibitive. A more modest though much 
less effective proposal would be to establish at least 
one meridional cross section from the Arctic to the 
Antarctic. 
There is also certain additional work which can be 
accomplished with our present material which could 
assist in a better definition of the character of the 
general circulation. It must be remembered that the 
normal state of the general circulation is not something 
which can be determined once and for all. It is some- 
thing toward which successive approximations must 
be made. In the Northern Hemisphere the rate of in- 
crease in aerological data over the past ten years has 
been large. A stock-taking of our present rate of accu- 
mulation of this material indicates that new monthly 
normal charts for upper levels should again be prepared. 
This might be especially worth while in polar and sub- 
tropical regions. It would appear that the present rate 
of accumulation of aerological information would make 
necessary the routine preparation of upper-level nor- 
mals at least every five years. 
Physical Climatology of the General Circulation. The 
large-scale pressure patterns and other features revealed 
by a study of the normal or average conditions of the 
general circulation are generally assumed to have defi- 
nite physical significance in the sense that they repre- 
sent a stable stationary state of the circulation rather 
than a haphazard combination of daily highs and lows. 
To the extent that this is true it is desirable to find 
logical physical-mathematical definitions and explana- 
tions of these normal features, since they represent 
relatively simple states of the circulation which up to 
the present have been easier to depict than individual 
daily circulations. 
A striking feature of the normal charts for the North- 
ern Hemisphere [30] are the very long waves of finite 
amplitude at upper levels which are superimposed on 
the circumpolar westerly current. While their amplitude 
increases with latitude and decreases with altitude in 
the troposphere, their wave lengths do not change 
markedly with either of these coordinates. 
These waves have dimensions greater than those of 
any other atmospheric wave systems. The smallest of 
them has a length of about 4000 mi and an amplitude 
of about 550 mi, compared to a dimension of about 
1000 mi for waves associated with extratropical cy- 
clones. Other interesting features are the consistently 
larger wave lengths and amplitudes found over Asia as 
compared to those found over North America, and the 
marked seasonal variations. 
Some of these features appear to be explained at least 
qualitatively by the simple theory of conservation of 
vorticity as applied to planetary waves. Thus Rossby 
in 1939 [44] was the first to explain on a theoretical 
basis how a system of waves could be superimposed on 
the circumpolar band of westerlies, with troughs and 
ridges located in fixed geographical regions. He ex- 
plained the anchoring of certain waves on the basis of 
distortions of the westerly flow due to north-south ori- 
