AEROLOGY OF TROPICAL STORMS 
By HERBERT RIEHL 
University of Chicago 
INTRODUCTION 
One of the earliest controversies that arose in regard 
to tropical storms concerned their vertical extent. Some 
writers claimed that disturbances disappeared at heights 
as low as 3 km—that people standing on high mountain 
tops could look down on the violent cyclonic whirls 
beneath them. They believed that the cirrus motion at 
upper levels remained undisturbed and indicated the 
direction in which the cyclone was moving. Others in- 
sisted that hurricanes must reach to great heights, 
10 km or more. They thought that cirrus radiated in all 
directions from the storms and used the change of direc- 
tion of cirrus movement with time to locate position 
and track of centers. 
The answer to this argument is now well known. As 
shown theoretically by Haurwitz [17] and empirically 
by high-level observations, mature storms extend 
through the troposphere. This affects such practical 
problems as routing of aircraft and forecasting storm 
movement. Since storms extend to the high tropos- 
phere, the upper layers should in part control the move- 
ment. We cannot hope to find the solution solely on 
sea-level or low-tropospheric charts, which very often 
do not reveal the state of the upper levels. 
As we ascend from the surface to the higher 
atmosphere, the simple trade-wind current and the clear 
demarkation line separating tropics and middle lati- 
tudes disappear. Above 500-400 mb we encounter a 
chain of complex vortices of large dimension that are 
interlocked with the troughs and ridges of the long 
waves in the westerlies [27, 29]. Interaction between 
low- and high-latitude disturbances largely controls 
short-term developments in the tropical vortex train 
aloft and affects the motion of typhoons and hurricanes. 
It also plays a decisive role in the events that lead up 
to storm generation. 
Formerly, formation and movement of hurricanes was 
forecast with detailed local analysis in a small area of 
the tropics. Since we know today that external forces 
are a major variable, we must expand the map analysis 
to include wide regions of the globe at all latitudes. 
Nevertheless the possibilities of detailed analysis are 
by no means exhausted. A complete three-dimensional 
description of hurricanes, such as has been rendered for 
middle-latitude cyclones, is as yet in the distant future. 
In the following pages we shall consider the state of 
knowledge regarding the structure, formation, and 
movement of tropical cyclones. Since all knowledge 
depends on the observations and their evaluation, we 
shall begin with tropical data and methods of analysis. 
TROPICAL DATA AND METHODS OF ANALYSIS 
Observations. Tropical storms form and move over 
water. Therefore the number of observations available 
for analysis and forecasting have always been notori- 
ously scarce. Even if perfect forecast methods were 
developed, the actual prediction would largely remain 
a guessing game for this reason. Apart from persistence 
and statistics on mean paths, even the best forecast 
method cannot help if there are no observations. 
The data useful for three-dimensional analysis are 
cloud observations, and commercial-aircraft, pibal, ra- 
win, and raob reports. Cloud observations and reports 
from commercial aircraft are most frequent, least costly, 
and also least useful. Wind observations from aireraft 
help only if pilots take double or triple drift measure- 
ments. This seldom happens. A good description of the 
state of the sky in surface observations still is hampered 
by the archaic code. Writers have pomted out the 
variety of cumuliform clouds that exist and have shown 
that broad inferences can be drawn from knowledge of 
the field distribution of cumulus types; but the code 
admits only the conventional L,, L,, and ZL; clouds. 
Many forecasters have also complained that the 
quality of cloud reports has gone down with the advent 
of sounding balloons. This holds particularly for direc- 
tion of middle- and high-cloud drift. At the beginning 
of this century the quality of these observations was 
high and the nephoscope was in frequent use. In those 
early days some of the articles which appeared were 
based on information about wind currents at high levels 
which we cannot match today. No wonder that some 
of the deductions rose far above the general level of the 
day and that they still form our major source of knowl- 
edge on upper-air currents over many areas, notably the 
tropical Atlantic Ocean. 
It is unreasonable that the tropical forecaster should 
be deprived of the potentially largest source of informa- 
tion and have to rely on the few stations with expensive 
sounding-equipment. The writer proposes the following 
action: 
1. Revision of the surface code to permit inclusion 
of at least three types of cumulus in one message. 
(There are about eight basic types that should be 
entered in the manuals on surface observations.) 
2. Insistence on careful estimates of altitude and 
motion of middle and high clouds. (The code should 
allow for transmission of cloud direction in two layers.) 
3. Installation of nephoscopes at all stations that 
make six-hourly reports, on naval vessels, and on re- 
liable commercial ships on the principal shipping lanes. 
Instrumental Measurements. Up to the present, the 
pilot-balloon observation has enjoyed the widest use 
among the instrumental measurements, but its draw- 
backs are numerous and well known. Balloons seldom 
reach the high troposphere. They are lost in clouds. 
In bad-weather areas where information is most needed, 
902 
eed 
