AEROLOGY OF TROPICAL STORMS 
position have been explored, while the dynamics have 
been neglected. It is certain that most tropical de- 
pressions form in consequence of superposition. Yet the 
great majority of these circulations do not develop 
beyond a weak wind field with maximum speeds of 
about Beaufort 6. Sometimes such weak centers will 
persist in steady state up to a week and then suddenly 
deepen. 
The Hypothesis of J. S. Sawyer. A major synoptic 
problem, therefore, remains: Under what circumstances 
will storms attain great intensity? The solution of this 
question must be connected with the mechanism of the 
upper outflow. Durst and Sutcliffe [13] have given a 
plausible explanation for the outflow in case of the 
mature hurricane. Their reasoning, however, is not ap- 
plicable to the period of establishment of a storm. 
Sawyer [385] adopts a line of thought that has also been 
put forward to explain cyclone formation in higher lati- 
tudes.1 The basic principle is given by Solberg [40], who 
showed that a ring of air particles may become “dy- 
namically unstable” provided that 
jars < ©, 
where f is the Coriolis parameter and ¢ the relative 
vorticity about the axis normal to the earth’s surface. 
For an explanation of the criterion the reader is referred 
to the articles quoted. 
Although Sawyer’s approach is novel and attractive, 
there are serious objections. In particular, it is dubious 
whether zones of “dynamic instability’? exist in the 
tropies prior to storm development. It is true that the 
criterion is much more easily realized in low than in 
high latitudes, since the value of the Coriolis parameter 
is much smaller. But a plausible mechanism for pro- 
duction and maintenance of negative absolute vorticity 
has not been given, nor has its actual existence in 
tropical currents prior to cyclogenesis been demon- 
strated. Within the writer’s experience, a zone of strong 
anticyclonic shear, comparable to that on the south 
side of the cireumpolar westerlies, does not exist in the 
tropics, at least within the latitude belt where tropical 
storms form. Sawyer himself [36] appears to be a little 
worried about his 1947 statements. 
The Hypothesis of H. Riehl. The attempt by Riehl 
[30] to explain the increase in intensity, does not postu- 
late any special internal characteristics within the zone 
of cyclone formation. It is based on the introduction of 
external forces aloft to produce the initial upper diver- 
gence. A number of observational facts are utilized for 
depicting the setting in which certain types of storms— 
but not all—develop: 
1. Deepening takes place when the curvature of the 
high-level flow is anticyclonic, or changes from cyclonic 
to anticyclonic. A high-level anticyclone or wedge is 
situated near (commonly west of) the surface disturb- 
ance, and an upper-level cyclone or trough lies to its 
east (Fig. 4). 
2. Superposition of these high-level centers and the 
1. Consult “Extratropical Cyclones” by J. Bjerknes, pp. 
577-598 in this Compendium. 
909 
wave troughs and ridges in the polar westerlies occurs 
so that high- and low-latitude systems are in phase. 
This strengthens the meridional flow components aloft 
and weakens the zonal component (Fig. 5). The in- 
tensification of the current above the surface depression 
must be brought about by acceleration toward lower 
pressure, therefore toward the east in the model chosen 
(Fig. 4). This eastward acceleration will give an initial 
pressure fall at the surface in the tropical disturbance. 
Fie. 4——Model of streamlines and contours (dashed lines) 
at 200 mb during deepening of a tropical disturbance (heavy 
dot) in the Northern Hemisphere. 
3. There is a general equatorward advance of polar 
air over the ocean east of the depression. This polar air 
does not enter the circulation. It subsides and diverges,” 
as usual, on its way toward lower latitudes. Continuity 
demands upper inflow above the region of subsidence. 
This further contributes to eastward deflection of the 
Fig. 5.—Streamlines of perturbation centers at 200 mb dur- 
ing in-phase (left) and out-of-phase (right) superposition of 
a trough in the westerlies on disturbances in the tropical vor- 
tex train aloft. 
high tropospheric air toward the east. Convergence and 
descent take place at the left edge of the upper current, 
looking downstream (Fig. 6). Divergence aloft develops 
at the right edge. It produces a surface pressure fall, 
2. Such divergence can also be obtained by ‘‘instability of 
the trades’’ [27]. 
