TROPICAL CYCLONES 
tainly it appears that tropical cyclones will develop 
only over comparatively warm water, probably 82F or 
higher. In 1890 only one tropical storm was noted in 
the North Atlantic and in several other years the total 
was only two. Therefore, it is apparent that a very 
special set of circumstances is necessary for their de- 
velopment. There is as yet no generally accepted defi- 
nition of exactly what synoptic situation is responsible 
for the formation of a tropical cyclone. 
In the North Atlantic, Riehl (12] found that intensifi- 
cation of low-level tropical perturbations or a change 
from a stable to an unstable condition resulted from 
the superposition of high-level extratropical disturb- 
ances. In all cases of hurricane formation noted in the 
course of this study, deepening began, without excep- 
tion, in pre-existing tropical disturbances. 
WEST O EAST 
895 
Cape Verde Island region—August and Septem- 
er. (It may be that most of these storms do not 
reach hurricane intensity until they reach longitude 
50-60°W.) 
Northern Caribbean Sea—late May through 
November. 
Just to the east and north of the West Indies— 
June through October. 
Southwestern Caribbean Sea—principally June 
and October. 
Gulf of Mexico—June through October. 
2. North Pacific Ocean off the west coast of Mexico— 
June through November. 
3. North Pacific Ocean, longitude 170° westward in- 
cluding the Philippines and the China Sea—May through 
December and (rarely) other months. 
Fig. 7.—Areas of tropical cyclone development. 
Tropical cyclones in the southwestern North Pacific 
Ocean for the year 1945, and to some extent for 1946 
and 1947, have been analyzed by Riehl [13] and several 
other members of the Institute of Meteorology, Uni- 
versity of Chicago. Riehl found that typhoon develop- 
ment can begin in consequence of instability of the 
Northern Hemisphere trades, not necessarily as a re- 
sult of interaction between currents of the two hemi- 
spheres. Instability of the trades appeared to set in 
just as a low-level wave was bypassed by a broad- 
scale ridge in the high troposphere. Riehl suggests 
that a tropical disturbance will intensify when moving 
from an upper trough to a high-level ridge, since this 
pressure distribution facilitates upper outflow from the 
storm. So far observational data have been insufficient 
to test these relationships in the tropical North 
Atlantic. 
An intensification of the observational network which 
will permit accurate trough and ridge analyses between 
the equator and latitude 30° during the tropical cyclone 
season is badly needed for further research on the 
interrelationship between upper troposphere troughs 
and ridges and the development of tropical cyclones. 
Region of Origin and Season. Tropical cyclones de- 
velop in the following areas and seasons: 
1. South portion of the.North Atlantic Ocean. 
4. North Indian Ocean. 
Bay of Bengal—April through December. 
Arabian Sea—April through June, September 
through December. 
5. South Indian Ocean. 
Madagascar eastward to 90°H—November through 
May. 
Off northwest coast of Australia—November 
through April. 
6. South Pacific Ocean from east of Australia east- 
ward to 140°W—December through April. 
Probably no portion of the tropical oceanic area 
in either hemisphere is entirely free from tropical 
storms, but there is no record of a true tropical cyclone 
of hurricane intensity in the South Atlantic Ocean or 
in the South Pacific east of about longitude 140°W. 
During the Southern Hemisphere summer, the inter- 
tropical front in these areas moves only a degree or so 
south of the equator, not far enough for the Coriolis 
force to become effective. Few, if any, tropical cyclones 
have been observed within 5 degrees of the equator. 
In Fig. 7 are shown the principal areas of the world 
where tropical cyclones develop and are observed. 
Frequencies of Tropical Cyclones. Data on the fre- 
quencies of tropical cyclones by months are summar- 
ized in Table I. 
