774 WEATHER FORECASTING 
colder half, and favor the existence of highs in the 
confluent colder half and the difluent warmer half. 
In the rear of a marked cold front of a more or less 
convex shape as seen from the warm-air side, a large 
cold-air outbreak, moving equatorward and invading 
regions of warmer air, spreads at the surface of the 
earth, a process which is connected with the produc- 
tion of anticyclonic vorticity in lower strata and the 
sinking of the center of gravity of the whole system of 
cold and adjoming air. It is obvious that such anti- 
eyclogenesis will be more efficient, all other factors 
remaining the same, the greater the temperature dif- 
ference between the cold and the warm air and the 
more high-reaching and vast the cold outbreak. Al- 
though the real arctic air does not reach so high, 
synoptic experience seems to indicate that the out- 
breaks of arctic air often form more intense and stable 
anticyclones than polar air outbreaks. This may be 
explained by the fact that the whole process attending 
a pronounced outbreak of the arctic front equatorward 
leads to a much greater drop of temperature in tem- 
perate latitudes than a corresponding advance of the 
polar front. The arctic air and the overlying polar air 
in the former case have their origm im much higher 
latitudes. 
Anticyclogenesis depends upon the supply of kinetic 
energy made available to it by the juxtaposition of 
tropical air over the polar air in which the anticyclone 
has first been formed. The forecaster should be on the 
lookout for regions on the constant-pressure maps 
where a zonal upper westerly with very high velocity 
is bounded (to the north and to the south) by two 
regions with relatively low zonal velocity. Elliott [23] 
has pointed out that an intensifying belt of abnormally 
strong upper westerlies exists for several days prior 
to the anticyclogenesis, and reaches a maximum just 
before it. Associating this development with an in- 
creased north-south thermal gradient across the region, 
he suggests that such anticyclogenesis is associated 
with excessive heat exchanges across the normal wester- 
lies between tropical and polar air. There are some- 
times other clues which are helpful in anticipating 
anticyclogenesis. On the surface maps, there is often 
the development of low pressure, a few days prior to 
anticyclogenesis, at half the distance of a stationary 
long wave (7.e., about thirty degrees of latitude) up- 
stream or downstream. 
Using the quasi-static assumption, we see that the 
pressure change at the base of the atmosphere is asso- 
ciated with the depletion or accumulation of air aloft. 
In particular, the pressure tendency at the ground is 
generally decided as to sign by the horizontal (mass) 
divergence or convergence above the level of non- 
divergence (see the following paragraph for substan- 
tiating evidence). This divergence (or convergence), 
acting in the same sense through the stratosphere and 
the upper half of the troposphere, may thus be an 
important effect to consider in the development and 
movement of sea-level systems. The patterns of veloc- 
ity divergence are of course determined by the dis- 
tribution of the instantaneous ageostrophic upper flow. 
As will be shown in a following subsection on the 
prediction of winds, a relatively large indraft of the 
streamlines across the absolute isohypses occurs in the 
right half of the entrance region and large outdraft in 
the right half of the ext region (see Fig. 3a). On the 
1 
(b) 1|OOO-MB MAP 
-———— RELATIVE ISOHYPSES 
=——— ABSOLUTE ISOHYPSES 
—-—-— STREAMLINES 
INTERSECTION OF IOOO0-MB ISOHYPSES WITH THE 
RELATIVE ISOHYPSES 
====== POSITION OF THE FRONT AT THE UPPER MAP 
Fic. 3.—The propagation, deepening, and filling of sea-level 
pressure disturbances (b) as determined from the absolute 
hypsography (a) of an upper constant-pressure map, according 
to Ryd-Scherhag [71]. 
other hand a relatively small angle of geostrophic de- 
viation (indraft and outdraft, respectively) is found 
in the left half of the entrance and exit regions. Hence 
there is velocity divergence in the exit (or delta) re- 
gion, convergence in the entrance (or confluence) re- 
gion. Provided that the distribution of the horizontal 
mass divergence can also be represented by this ‘‘delta- 
divergence” and ‘‘confluence-convergence,” then the 
pressure tendency at the ground should be positive 
—— 
— 
