AEROLOGY OF TROPICAL STORMS 
trary to Deppermann, a gradual horizontal increase of 
temperature across the rain area, based largely on 
three-hourly special radiosonde observations at Miami, 
Florida, during the approach of a severe hurricane in 
September, 1947. Other series of special observations 
also fail to disclose a sharp boundary. 
Several explanations of the difference between these 
authors are possible. Of these the most plausible one is 
that conditions are not symmetrical around hurricanes. 
A “wall” may be present in one or two quadrants but 
not the others. There is another interesting possibility. 
As pointed out by Byers [6], the temperature of the 
surface air spiralling toward a center must fall 3C or 
more during horizontal motion because of the large 
pressure reduction. Such a temperature decrease is 
seldom, if ever, observed. The temperature difference 
between ocean and surface air suddenly increases with 
lowering air pressure. As the ocean is greatly agitated 
and as large amounts of water are thrown into the 
atmosphere in the form of spray, very favorable condi- 
tions exist for rapid transfer: of sensible and latent heat 
from ocean to air. Such transfer could account for a 
gradual increase of upper-level temperatures across the 
rain area. Figures la and 1b show that the effect referred 
to is very large. It roughly doubles the size of the 
positive area enclosed by inside and outside soundings 
on the thermodynamic diagram. 
Microstructure of the Rain Area. There is conclusive 
evidence that conditions are not uniform as we circle a 
hurricane. Deppermann [10] showed that barometric 
fluctuations of shorter period are superimposed on the 
general barometric trace. These fluctuations have differ- 
ent periods, varying from minutes to hours. Depper- 
mann [10] associated them in part with variations in 
cloud thickness and rain intensity. 
The existence of a nonuniform cloud structure was 
proved by radar photographs of cloud systems in hurri- 
canes [43]. On these photographs we can see that the 
rain area is composed of a number of bands of heavy 
precipitation that alternate with more settled condi- 
tions. The bands spiral cyclonically toward the center. 
Photographs of this kind have since been taken in large 
numbers. Further descriptive and analytical discussion 
of this aspect of the microstructure is one of the more 
interesting topics that await the investigator. A possible 
explanation is the development of internal waves. 
Another is the effect of falling rain and vertical down- 
drafts as discussed by Byers and Braham [7]. An 
analogy between the radar bands of Wexler and the 
thunderstorm-cell propagation proposed by Byers and 
Braham is by no means illogical. 
Thermal Structure of the Eye. The eye of the hurri- 
cane has held the attention of all who have written on 
hurricanes from the earliest days. It is one of the oddest 
curiosities in meteorology. The precipitation ceases 
abruptly at the boundary of well-developed eyes, the 
skies clear, and the wind subsides suddenly. Observers 
inferred early that the vertical motion in the eye should 
be downward, the air therefore very warm and dry. 
Some thought that the tropopause was “sucked down.” 
Aerological observations have confirmed a part of 
905 
these inferences. Pilots flying into eyes at various alti- 
tudes have reported large and sudden temperature 
increases. The best evidence comes from two radio- 
sonde flights taken at Tampa, Florida, in two hurri- 
canes [28, 39]. It is, of course, very difficult to send up 
balloons in an eye. In particular, there is no certainty 
that the balloons were still in the eye when they reached 
the high troposphere. In both cases, however, the eye 
was large. Chances are good that the balloons did not 
drift outside and that the temperatures reported at all 
levels are really indicative of the structure of vertical 
columns. 
The sounding of October, 1945 [28] (Fig. 1a) showed 
great stability in the lower troposphere and a very 
steep lapse rate higher up. From the 100-mb level 
downward, temperatures were higher than otherwise 
observed in the tropics. Departures from the normal 
tropical atmosphere were extreme from 800 to 400 mb. 
They indicated subsidence amounting to several kilo- 
meters. The two Tampa soundings disagreed as to 
temperature in the high troposphere. In the case de- 
scribed by Simpson [39] the air above 300 mb was 
colder than that outside the eye. This observation is 
very difficult to interpret. The October 1945 ascent 
showed equalization of temperature with the surround- 
ings only near 100 mb. Above this level, the air prob- 
ably was colder inside the eye, so that the entire thermal 
structure relative to the outside resembled that of 
dynamic highs. Neither sounding reached the tropo- 
pause, and there was no evidence of “sucking down.” 
On the contrary, it is probable that the tropopause is 
highest above the eye, again in analogy to warm anti- 
cyclones. 
The extremely warm air of the eye is necessary to 
explain the very low surface pressures observed, at least 
statically. Haurwitz [17] has shown that application of 
the hydrostatic equation is entirely valid for computa- 
tions of the vertical pressure distribution even in hurri- 
canes. We can calculate, for example, what reduction 
of sea-level pressure is possible if the normal tropical 
atmosphere is replaced by the atmosphere of the rain 
area and if the pressure remains constant at the level 
where the two soundings intersect. This reduction is of 
the order of 30 mb and may reach 40 mb on the outside. 
It does not suffice to explain the very low sea-level 
pressures observed in many cases (900-950 mb). If we 
introduce an eye with a sloping boundary, however, 
following Haurwitz [17] and others (Fig. 2), the problem 
is resolved. Introduction of air still warmer than that 
of the rain area makes possible additional large pressure 
reductions. 
Broadly speaking, the thermal structure of the eye, 
at least below 300 mb, is relatively well established. 
Many observations, of course, are needed to give all 
details, especially the slope of the boundary at different 
heights. Such observations would be of great scientific 
interest, but probably of little practical value. 
Dynamical Aspects of the Rain Area. The thermal 
structure of the rain area permits some direct inferences 
regarding the dynamical structure of hurricanes. We 
have seen that soundings in the rain area coincide with 
