612 
facts concerning the movement of cirrus clouds as, for 
example, shown by Pick and Bowering [49] im 1929. 
Sawyer [1] has shown in an interesting study that the 
change of the vorticity field at different levels during 
the deepening of a cyclone can be used for computation 
of the vertical movement of the air. Instead of the real 
vorticity field, which is difficult to determine, Sawyer 
used the vorticity field computed from the pressure 
Fic. 7—Schematic diagrams of young wave cyclone and 
mature occluded cyclone. Solid thin lines are sea-level isobars, 
dashed lines are 500-mb streamlines. Precipitation indicated by 
stippling. 
field. Since the pressure field gives an approximate value 
of the wind field, the change of vorticity can be com- 
puted from the pressure changes on consecutive maps. 
The method is essentially the same as that used here in 
order to describe the formation of the characteristic 
upper contour Imes in an occluded cyclone. 
FORMATION OF HIGH-LEVEL CYCLONES 
AND ANTICYCLONES 
It was pointed out earlier that the cold upper troughs 
undergo continuous changes and that they have a life 
history of growth and ultimate degeneration as have 
the surface cyclones. In some cases the upper troughs 
seem to be disturbances of larger scale than cyclones; 
in other cases, however, the dominating upper-air flow 
MECHANICS OF PRESSURE SYSTEMS 
is from the west and every individual migrating upper 
disturbance is associated with a surface cyclone. 
In the following paragraphs some typical cases de- 
seribed in the schematic chart in Fig. 4 will be dis- 
cussed in more detail. As an example of the develop- 
ment of a strong upper-level low from a pre-existing 
trough we select the case of November 17-19, 1948 
over North America. This situation also illustrates the 
characteristic structure of an occluded cyclone pre- 
viously discussed. The sequence of weather maps shows 
that the structure characteristic of the occluded frontal 
cyclone can be the result of processes other than the 
occlusion of wave-shaped frontal perturbations of the 
type shown in Fig. 7. 
Figure 8 presents five charts for the 500-mb surface. 
The contour lines and the schematic front show the 
rapid development of a deep upper cyclone from a rela- 
tively weak trough. On the chart for November 17, 
1945 the upper flow is predomimantly westerly with a 
cold trough aloft over the western part of the United 
States. The cold air mass grows gradually out to the 
south and at the same time a closed upper circulation 
develops. During the process, which can be followed on 
the five charts with a time interval of 12 hours, the cold 
air mass at the 500-mb level is gradually cut off from 
its polar source region, and on the chart for November 
19, 1500 GMT there is left a cold “drop” separated from 
the cold masses in the north. 
The thermal structure of the air masses can be studied 
in the vertical cross section (Fig. 9) along the broken 
line marked on the chart for November 18, 1500 GMT. 
Especially along the west side of the upper trough, the 
sloping front is very well marked. The wind component 
normal to the cross sections shows the characteristic 
concentration of the maximum wind velocity in a pro- 
nounced jet. Figure 9 thus shows that the upper trough 
consists of a central part filled with cold air masses of 
polar origin surrounded by warmer air masses. The low 
tropopause and high stratospheric temperature and the 
multiple tropopause are very characteristic of all similar 
situations. The fast-moving upper air mn the region of 
the lower stratosphere is subjected to a strong subsi- 
dence on the west side of the trough and to a strong 
ascent on the east side (see, for example, [24]). 
In Fig. 10 the contours of the principal front at the 
surface and on the 850-, 700-, and 500-mb surfaces are 
presented for November 18 and 19, 0300 and 1500 
GMT. At the beginning of the period there are two 
surface fronts, one western front connected with the 
developing disturbance and one eastern front corre- 
sponding to the northern boundary of the moist warm 
Gulf air. On the chart for November 18, 1500 GMT the 
surface fronts have already been brought so near each 
other that a separation is impossible. At that time a 
structure corresponding to the classical picture of an 
occluded cyclone has already started to develop. On the 
next day the structure of a regularly occluded cyclone 
is complete, as can be seen from the surface map in 
Fig. 11. 
In Fig. 8d the principal precipitation areas are 
marked. From the figure it can be seen that the areas of 
