900 
ricane. Waves generated in the right half of the storm 
travel with the storm, so they are under the influence of 
winds with relatively little deviation in direction for a 
longer time, 2.¢., they have a comparatively. long fetch. 
Thus the strongest swells, which move through the © 
storm and eventually a long distance ahead of the 
storm, are produced in this half. Conversely, swells 
traveling opposite to the direction of movement of the 
storm will be weaker. (See Fig. 9.) 
It has been stated earlier that swells emanating from 
the region of a tropical cyclone are useful in its early 
detection and location and in the determination of its 
intensity. Swells also have further forecast value. If 
the direction of movement of the swells remains the 
same, the storm is either approaching directly or going 
away from the observer. If the direction changes coun- 
terclockwise, the storm is passing from the observer’s 
right to his left; if it changes clockwise, it has passed 
or is passing from his left to his right, as he faces the 
swell. The direction from which the swell approaches 
indicates the direction of the storm from the observer 
at the time the swell was generated. 
Forecasting Significance of Storm Tides. Particularly 
along the shore lines of partially enclosed bodies of 
water, such as the Gulf of Mexico, tides have some 
forecast value. Abnormally high tides are indicative 
of the presence of a storm. An area along the coast line 
where the tide exceeds the predicted gravitational tide 
and continues rising is in lme with the advance of the 
tropical storm at the time the water started on its 
journey. If the point of greatest plus departure from 
the predicted tide shifts right or left, this indicates 
that the storm is changing direction toward the point 
where the increased rise is taking place. 
Microseisms. Microseisms are more or less regular 
elastic surface waves which may result from a number 
of causes, including ocean waves and surf. Microseisms 
of these two types may be propagated over consider- 
able distances, if no deep geological discontinuities 
exist, and may be measured at seismographic stations. 
Gutenberg [8] has pointed out that the location of 
tropical cyclones, and their intensities, direction, and 
rate of movement, can be determined from the differ- 
ences in arrival time at three stations on a triangle, by 
plotting relative amplitudes and considering the 
changes with time. It appears that this technique might 
have its greatest application where airplane recon- 
naissance and ship reports are not available. 
NEEDED RESEARCH ON TROPICAL 
CYCLONES 
Additional Data Required. Since tropical cyclones 
usually develop in and often traverse remote oceanic 
areas, the acquisition of data required for a rigorous 
evaluation of many of the current theories of structure 
and development of these storms and for a solution of 
urgent forecast problems has been difficult or impos- 
sible. This is especially true with respect to upper-air 
information. Since tropical cyclones are relatively small, 
a denser network of observing stations is a prerequisite 
to further progress. Over large oceanic areas upper-air 
TROPICAL METEOROLOGY 
data have been negligible. Southern and western 
Florida, the most hurricane-susceptible area in the 
United States, would provide an excellent laboratory 
for hurricane study if a satisfactory network of ob- 
serving stations could be set up. This network should 
be supplemented by a fleet of “‘weather-trucks” and 
mobile weather stations which could be dispatched 
and located to form any desired observing grid. Over 
adjacent water areas squadrons of planes, equipped 
with the latest meteorological devices for measuring 
winds, convergence, divergence, updrafts and down- 
drafts, should be dispatched into and over the storm 
area for inflight and parachute-radiosonde observations. 
Photographs over the top of the storm and in the eye 
should be taken for study purposes. For a period of at 
least two years, during the hurricane season, stationary 
ships should be placed at 300-500 mile intervals in the 
Gulf of Mexico, in the Caribbean Sea, and between 
latitudes 8°N and 30°N in the Atlantic for the purpose 
of securing regular surface and upper-air observations. 
Urgent Research Needs for Forecasting. Antecedent 
Prerequisites for Tropical Cyclone Development. The exact 
combination of conditions which produces tropical cy- 
clones must occur rarely, since the phenomenon itself is 
infrequent. The original causative factors which change 
the normally very weak and transitory pressure and 
wind deformations in the trades and the intertropical 
convergence zone to unstable waves and from the latter 
to intense tropical storms are not fully known. The 
testing of Riehl’s recently published ideas on tropical 
cyclogenesis based on the interaction of surface per- 
turbations and upper troughs and ridges should have 
very high priority as a research project. Radiosonde 
data, preferably from stationary ships, would be re- 
quired. In the event the currently most persuasive 
theories are not substantiated, the new data would 
prove invaluable in deriving new ideas and further 
clues on the origin of tropical cyclones. 
Differences in the intensities of tropical cyclones, 
which may be of the order of 80 to 90 mb, appear to 
depend upon apparently slight differences in the mois- 
ture and stability of inflowing air during the stage of 
marked development, with little change thereafter. 
Practicable techniques of adequately measuring and 
determining these differences and the establishment of 
criteria for limiting pressure gradients and wind shears 
in the developing storm would provide valuable tools 
for the forecaster. 
Effective Steering Level. Although forecasters are using 
certain techniques based on steering at some arbitrary 
level or immediately above the s.orm circulation, 
further objective tests of these empirical rules are 
desirable. Recurvature is the most difficult problem 
facing the forecaster and, while lack of adequate ob- 
servational data may always be a problem in certain 
areas, more definitive techniques seem possible. 
Research Needed for More Complete Knowledge of 
Tropical-Cyclone Structure. 1. Tropical meteorology is 
deeply in debt to Father Deppermann for his pains- 
taking assembly and analysis of typhoon characteristics 
in the Philippine area, but a further systematic collec- 
