University of Florida
Florida State Planning Board
Engineering Experiment Station
Bulletin No. 3
t he locating of Cropical Storms
by Sleans of cTissociated Static
JOSEPH WEIL, M.S.
Professor of Electrical Engineering
University of Florida
State and University Radio Station WRUF
University of Florida
Plans for the experimental work described in this bulletin
were discussed on June 15, 1935, in a conference in Wash-
ington attended by Dr. Gleason W. Kenrick and Dr. Carlos
Chardon of the University of Puerto Rico. and Professor
Joseph Weil of the University of Florida.
Through the efforts of Mr. Josiah Ferris, Secretary to the
Senate Committee on Naval Affairs. the United States
Weather Bureau secured from the United States Navy
Department the loan of equipment which formed the basic
apparatus for the research described in this bulletin.
The writers are particularly indebted to Dr. Kenrick for
his cooperation. Many of the changes made in the ap-
paratus and technique were the result of suggestions offered
by him during the almost daily discussions on the progress
of the experiments by means of short wave radio.
The staff of radio station WRUF. particularly Mr. F.
Banks Duncan, remodelled the original Watson-Watt equip-
ment and materially assisted in the design and construction
of the photographic and recording apparatus.
Necessary funds were obtained from the Works Progress
Administration. Particular mention should be made of the
close cooperation given this research problem by Mr. E. A.
Pynchon. Florida WPA Administrator and Mr. William L.
Wilson, Research Coordinator, and their associates.
Members of the staff of the Florida State Mapping Pro-
ject. particularly Mr. Malcolm Bruce. were of much aid in
preparing the necessary maps.
Others to whom credit is due include Mr. Silas Thronson
and Mr. Ralph Sneeringer, who did much of the photo-
graphic work: Mr. F. HI. Hayes, who constructed the photo-
graphic equipment: Messrs. H. V. Thompson and John P.
Lenkerd, who assisted in the construction of the trans-
mitter; Messrs. William B. Bernard and William A. Hicks,
who made many of the observations and assisted in com-
piling the results.
Professor S. P. Sashoff, of the Department of Electrical
Engineering of the University, has given valuable advice
in editing this bulletin.
The authors wish to particularly acknowledge the helpful
assistance received from Dean Blake R. Van Leer of the
College of Engineering. They also desire to express their
appreciation to the Florida State Planning Board for their
assistance in making possible the printing of this bulletin.
The chief purpose of an Engineering Experiment Station
is to seek new truths, and to make discoveries of value for
the industries and people of the State.
In the realm of research, the answers are not known-
if they were, there would be no purpose to the investiga-
tions. Frequently, therefore, the research worker after
painstaking efforts has only a negative result to repay his
In the case of this investigation-the third to be published
under the auspices of the Engineering Experiment Sta-
tion. and the second in cooperation with the Florida State
Planning Board-unusual good fortune has attended it.
It is fortunate that at the University of Florida, in the
College of Engineering, and in the staff of the State Radio
Station WRUF. there were men who were capable of con-
ducting such an investigation. It is fortunate that so many
Federally sponsored agencies could and would cooperate in
the investigation. It is fortunate that the technical in-
vestigations so quickly pointed to practical results. It is
fortunate that the Florida State Planning Board was will-
ing and able to assist in making these results available to
all who have need for them.
Hurricanes cause much worry to many inhabitants of
various parts of the world. Every geographical location on
the earth carries with it serious and frequently difficult
problems which require study, observation, synthesis and
analysis. Problems which will have to be solved by those
who have the talents, and who love their country and are
determined to preserve and use its great natural resources.
"The Location of Tropical Storms by Means of Associated
Static" is probably the first scientific attack upon this
problem by Florida men. This bulletin does not claim that
the problem has been solved. It merely gives a scientific
method of attack which, thus far, gives hope for practical
results. and one which should be tested and investigated to
the fullest extent.
This is but one of many such problems which could and
should be studied and reported upon. Other states are
spending hundreds of thousands of dollars trying to solve
their difficult scientific problems. The State of Florida has
not as yet made a single direct appropriation in aid of its
Engineering Experiment Station.
It is hoped that this bulletin will convince the people of
the State that the support of such investigations are worth
many times what they cost.
M. L. MONTGOMERY BLAKE R. VAN LEER
Executive Secretary Director
Florida State Planning Board Engineering Experiment Station
4 ENGINEERING EXPERIMENT STATION RB'LLETIN
The static crashes associated with a hurricane passing
over Florida in 1928 were observed by experimenters at
Houlton, Maine. Even five years prior to this, other ob-
servers at the National Physical Laboratories in England
had been able to secure good correlation between readings
obtained by an oscillograph and the azimuth of lightning
flashes near enough to be visible. Since 1928 other work-
ers in Canada, Australia and the United States have made
improvements on the apparatus used in the early experi-
mental work. However, insofar as is known, none of the
developments have produced apparatus that has proved sat-
isfactory for quick, accurate weather forecasting, particu-
larly with reference to hurricanes.
It was toward this end and with the idea of further im-
proving apparatus and devising a suitable technique of ob-
servation that stations were set up during the past year
at the University of Florida and the University of Puerto
Rico. The equipment used, together with observations se-
cured on a hurricane passing within two hundred miles of
the observing station at Gainesville, are discussed in this
The theory of operation of this equipment is based upon
the determination of the direction of atmospherics (static)
emanating from storms. Whether all storms send out such
radiation is not definitely known. Certainly, from visible
lightning, there will emanate an periodic discharge. Wheth-
er similar discharges are produced in connection with all
tropical disturbances remains to be determined. Nor can it be
said that visible lightning is the sole source of atmospherics
which can be detected by the radio goniometer. Therefore,
it seemed advisable to determine whether radiations, which
could be detected, were associated with tropical disturb-
ances. The data shown herein indicate that from directions
corresponding to the location of a hurricane such radiations
are obtained. It cannot be said that the static from the
hurricane predominates in all cases. Local storms fre-
TIHE LO('CAT7ING OF TROP'I('AL STORMS
quently cause far greater field intensities at the observing
station, and during parts of the day completely mask the
rays apparently emanating from the hurricane. However,
even though such local static might in some cases be quite
pronounced, in almost every case static may be found com-
ing, apparently, from the hurricane area.
Knowing the approximate position of the storm with
reference to the observing station, it should be possible to
select radiations coming, apparently, from that general di-
rection. From similar observations made at other stations,
the probable position of the storm might then be obtained
by the method of triangulation. The location of the prob-
able source of each individual static crash associated with
the storm might also be determined. Furthermore, from
such data several storms might be shown to occur simultane-
ously. In general, however, only one will remain in the
approximate direction of the hurricane for any appreciable
time, since tropical disturbances move at slower speeds than
do most other storms. From such data obtained at several
stations, information pertaining to the location of the hurri-
cane may be secured and used in conjunction with that ob-
tained by other methods and so make possible more accurate
The apparatus used in these experiments is based pri-
marily upon that developed by Watson-Watt in 1926, and
much of it was built under his direction for the United
States Navy Department. It was made available for our
experiments through the efforts of the United States
Weather Bureau and the United States Navy. It consists
essentially of two vertical loops, placed in planes mutually
perpendicular. Each loop is connected to an amplifier. The
output of each amplifier is then impressed across the deflect-
ing plates of a cathode-ray oscillograph, resulting in a fluor-
escent line, or deflection, appearing on the screen of the
tube when the loops are excited. Photographs of this
screen are taken by means of a specially constructed camera.
These are studied for correlations between the angles of
6 ENGINEERING EXPERIMENT STATION BULLETIN
Figure I.-Loop Aerials.
THE LOCATING OF1 TROI'ICAL STORMS
the deflections observed and the azimuths of received atmos-
pherics. A simplified diagram (Fig. 2) shows the general
arrangement of the apparatus. Details pertaining to the
individual units follow.
Figure 2.-Block Diagram of Radio Goniometer.
The loop aerials (Figs. 1 and 3) used in this experiment
consist of 32 turns of silicon-bronze wire wound on two
rectangular frames two meters to the side. A resistance
of two ohms. grounded at its midpoint, is inserted in the
electrical center of each loop (Fig. 3). These resistances
are necessary for the insertion of the calibrating electro-
motive force from the test oscillator.
A tuning unit consisting of a variable condenser with a
range of approximately 1.1 to 0.001 microfarads is provided
for adjustment of the loop aerials to resonance to any fre-
quency within the range of seven to twenty kilocycles per
second. The tuning unit is placed close to the amplifier and
associated equipment and about ten feet from the nearest
8 ENGINEERING EXPERIMENT STATION BULLETIN
THE LOCATING OF TROPICAL STORMS
portion of the loop aerials. The loop aerials are connected
to their tuning units through a transposed transmission line
in order to transfer the induced electromotive force from
the loops to the tuning units without too great an attenua-
tion and to reduce the effect of external fields.
A transformer, incorporating a Faraday screen, is used
for coupling the output of each loop to the grid circuit of
each amplifier. The Faraday screen is provided to eliminate
electro-static coupling (antenna effect) between the loop
and the amplifying equipment. The transformer was also
designed to provide efficient impedance matching between
the loop circuit and the grid circuit of the amplifier.
Figure 4.-Amplifier Wiring Diagram.
The resistance coupled amplifier as first used (Fig. 4)
contained British triode tubes, but these were replaced in
the modified amplifier (Fig. 5) with modern tubes of the 6F5
10 ENGINEERING EXPERIMENT STATION BULLETIN
and 6C5 types. This change not only reduced the difficulties
due to variations in the construction of the older type tubes,
but also reduced microphonic noises and hiss due to thermal
agitation, and resulted in an increase in gain from ninety
to one hundred and thirty decibels.
The gain of the amplifier can be changed in either of
two ways. Coarse gain adjustment is obtained by varying
a potentiometer between the second and third amplifier
tubes. Fine gain control is accomplished by varying the
grid bias to the first three stages of the amplifier. The out-
put of the final stage coupling element was changed from re-
sistance to impedance by removing a 100,000 ohm plate
resistor and replacing it with a 150 henry inductance, there-
by increasing the direct potential impressed upon the plate
of this tube. In order to further increase the electromotive
force applied to the deflecting plates of the oscillograph and
to afford more efficient matching, a 3/1 step-up transformer
is used for coupling the output of the amplifier to the
THE LOCATING OF TROPICAL STORMS
,'ir n- o t c^~^v
Figure 6.-Oscillograph Transformer Unit.
T- DAA oR, T /T u\
*''SSAC R^S/sra/VCCs ^ -
0, c 0 0- J
-L ~ or i -i ? r --
Ta T~< T~? 5\ I'^r
LT- G 2
*. / ^*
12 ENGINEERING EXPI'ERIMENT STATION BULLETIN
cathode-ray tube. A variable condenser is connected across
the secondary of this transformer for phase adjustment
(Fig. 6). Insofar as possible the amplifying systems used
between the loops and the deflecting plates of the oscillo-
graph unit are made similar.
Difficulties which arose due to coupling between the am-
plifier output lines and the loops necessitated placing these
lines in a grounded conduit. Particular attention was paid
to the elimination of outside influence affecting the loops
by mutual induction.
For purposes of alignment and adjustment of amplifier
gain, a shielded test oscillator (Fig. 7) is provided. The
oscillator is of the Hartley type. with one stage of resistance-
coupled amplification. The magnitude of the electromotive
force output can be varied by means of a potentiometer
control and the frequency is adjustable from 7.5 to 40 kilo-
cycles per second.
r- -------------------------- ------------1
L---- 8 f-------------------------J
Figure 8.-Buffer Unit.
TIE LOCATING OF TROPICAL STORMS
The output of this oscillator is transmitted over a con-
centric cable to a buffer unit (Fig. 8) and from there is
impressed across the one ohm resistance in the center of
each loop, terminals G-N' and G-E' of Fig. 3.
Figure 9.-496-A Power Supply Unit for Cathode-Ray Oscillograph.
Cathode-Ray Oscillograph Unit
The Cathode-ray Oscillograph unit consists of a five inch
R. C. A. 907 type tube having a General Radio Type 496-A
power supply unit. The original apparatus, shown in Fig.
9, was modified as shown in Fig. 10. This made it possible
to obtain independent variable voltage which could be ap-
plied to each set of deflecting plates for convenience in cen-
Figure 10.--(scillograph Power Supply.
14 E(NGINEERING EXPERIMENT STATION HI'LLETIN
tering the spot. These changes also raised the output poten-
tial from 1500 to 2000 volts, which was necessary for
satisfactory operation of the 907 tube.
A special camera, using 16 mm film, was constructed and
then mounted with the cathode-ray tube in a light-tight
housing. Two types of film driving mechanism were pro-
vided, one for making continuous or synchronous runs and
a second for taking composite exposures.
The composite exposure is obtained by exposing station-
ary film for periods of thirty seconds to the images formed
on the cathode-ray tube.
In making these exposures, a synchronous motor-driven
set of contacts closes the circuit for thirty seconds during
the first minute of each fifteen minute period of the hour.
These contacts operate a pair of relays contained in the
control panel. One relay makes instantaneous contact and
instantaneous break, and the other relay is similar except
that its break is delayed about two or three seconds. The
first relay operates the electromagnetic shutter of the cam-
era. which also moves the film one frame each time the
shutter is opened. The second relay applies anode voltage
to the cathode-ray tube and removes the charging voltage
from the filament battery. The filaments of the amplifier
tubes are kept hot during the entire observation period, but
the battery charging voltage is not applied while the ex-
posure is being made. When the two relays are de-energized
the camera shutter is closed two or three seconds before
the anode voltage is removed from the cathode-ray tube.
and the charging electromotive force is again applied to
the filament battery. By closing the shutter before de-
energizing the cathode-ray tube no record is made on the
film of the slow movement of the spot across the screen,
which occurs as the anode electromotive force across the
power supply condensers decreases.
Composite pictures are obtained automatically by relays
actuating the necessary mechanism. After the required
pictures are secured they are projected upon a compass rose
THE LOCATING OF TROPICAL STORMS
July 28 1936
7:45 PM 8:00 PM 8:15 PM
July 29 1936
7:45 AM 8:00 AM 8:15 AM
July 29 1936
7:45 PM 8:00 PM 8:15 PM
Figure 11.-Composite Exposures.
16 ENGINEERING EXPERIMENT STATION BULLETIN
July 30 1936
7:45 AM 8:00 AM 8:15 AM
July 30 1936
July 31 1936
7:45 AM 8:00 AM 8:15 AM
Figure 12.-Composite Exposures.
Figure 13.-Synchronous Exposures Taken at (ainesville (upper) and Puerto Rico (lower).
18 ENGINEERING EXPERIMENT STATION BULLETIN
so that the bearings of the static can be obtained directly.
The irregularly shaped object at the center of the rays
of the composites shown in Figs. 11 and 12 is a masking
spot. This is an opaque covering over the center of the
cathode-ray tube screen which prevents the central spot
from fogging the film.
Figure 14.-Recording Equipment.
For making the synchronized runs, of which Fig. 13
is an illustration, an additional mechanism, operated by
means of a synchronous motor, is provided so that the film
can be moved at a continuous speed of one inch per second.
In order to facilitate the comparison of simultaneous photo-
graphic records made from such continuous runs at more
than one station, special timing marks are placed on the
films. This is accomplished by sending coded signals from
radio transmitter W4XAD at Gainesville, Florida. The
signal at each receiving station is amplified and made to
operate a light source which is projected through a special
lens system upon the moving film, thereby placing upon each
film synchronous marks which assist in identifying asso-
ciated simultaneous crashes.
THEIk LOCA TING; OF TROPICAL, STOR.II
General Arrangement of Apparatus
A photograph of the apparatus at the recording building
in Gainesville. Florida, is shown in Fig. 14. In the left
foreground will be seen the two tuner units. two coupling
transformers (the vertical cylinders) and the rectangular
cases housing the two amplifiers. To the right of the am-
plifiers is the concentric cable through which the calibrating
electromotive force is fed from the test oscillator to the
Figure 15.-Receiving Position.
loops. To the right of the oscillator is the black metal hood
covering the camera and cathode-ray tube, the screen of
the latter being visible. To the right of the cathode-ray
tube is the unit containing the cathode-ray tube input trans-
formers and phasing condensers. The remaining unit is
the power supply for the cathode-ray tube. The synchron-
ous clock used in timing the composite exposures is placed
above the oscillator cabinet. The cathode-ray tube and cam-
era housing can be made light-tight, thereby permitting
operation of the apparatus even when the room is illum-
20 ENGINEERING EXPERIMENT STATION BULLETIN
The communication equipment, although not directly con-
nected with the recording of directional atmospherics, is
THE LOCATING OF TROPICAL STORMS
a very essential component of the project. Figures 15 and
16 show the receiving and transmitting equipment. This
equipment is housed about seven hundred feet east of the
recording building with all cables, power and communica-
tion, between these buildings carried underground. The
transmitting and receiving equipment, together with three
directional antenna systems, is capable of operation on any
one of several experimental frequencies ranging from 17
to 3 megacycles. In the near future it is expected to be
able to operate on two or more channels simultaneously. A
Bruce rhomboid antenna, directed toward Puerto Rico, is
used for reception of the radiotelephone signals of K4XAO.
After the loops have been placed in the proper geographic
north-south and east-west planes and tests and adjustments
have been made to determine that no mutual inductance
exists between the two loops, the amplifiers are aligned.
This is accomplished by placing the test oscillator in opera-
tion and adjusting its output until there is a noticeable
deflection on the screen of the cathode-ray tube. The tun-
ing units of each loop are then adjusted until a maximum
deflection is obtained. At this point the grids of the first
tubes of each amplifier are connected together through two
mica condensers to insure that the input potentials to the
two grid circuits are identical in phase and amplitude. With
the controls of both amplifiers set for approximately equal
gain, the deflection is noted. If the outputs of the two am-
plifiers are identical in amplitude and phase, there will ap-
pear on the screen of the tube a line making an angle of
45 degrees with the line secured by energizing either pair
of deflecting plates alone. If an ellipse having such a 45
degree axis appears, the outputs are identical as to ampli-
tude, but there is a phase displacement. This phase dis-
placement can be eliminated by adjustment of either or both
of the phasing control condensers which are connected in
the secondary circuits of the input transformers to the
cathode-ray tube deflecting plates (Fig. 6). or by compen-
sating adjustments of the coarse and fine gain controls of
22 ENGINEERING EXPERIMENT STATION BULLETIN
the amplifiers (Fig. 5). This latter method is possible be-
cause the coarse gain control, being a potentiometer placed
in the grid circuit of the third amplifier tube, causes a phase
shift as well as a variation of the potential applied to the
grid circuit, whereas the fine gain control, operating only
upon the grid bias voltage upon the first three tube grids,
will occasion little shift in phase although permitting a var-
iation of amplification. When the image upon the screen
is a straight line at 45 degrees, variations in the test os-
cillator frequency should cause merely a change in length
of the line without shifting the angle or forming an ellipse.
If such a change occurs, more accurate adjustment must
The connection between the grids of the input tubes is
now removed and the loop tuning units adjusted for a maxi-
mum deflection without changing the position of the 45
degree line or permitting it to become an ellipse. At this
point, it should be possible to connect and disconnect the
grids of the two input tubes without altering the angle
or changing the shape of the image. The only change
that might take place is a change in the length of the line.
If the oscillator is now turned off, rapid flashes at varied
angles can he observed upon the screen of the cathode-ray
tube due to received atmospherics.
The camera is focused with a blank piece of film in the
gate and then loaded with the shutter closed.
When synchronous runs are made, radiotelegraphic com-
munication is established through station W4XAD with the
collaborating station K4XAO at the University of Puerto
Rico. From Gainesville a coded radiotelegraphic signal is
transmitted. It is received at both Puerto Rico and Gaines-
ville recording stations, amplified and made to operate a
light source which is projected upon the edge of the film.
Thus. with the amplifiers and cathode-ray tube in operation,
the camera film stationary, and the shutter closed, the
operators at the recording laboratories await the pre-
arranged signal for starting the motor and opening the
shutter. After this, another signal is sent. This comprises
the transmission (in the Continental code) of the alphabet
THE LOCATING OF TROPICAL STORMS
at about ten words per minute, a ten second dash, the identi-
fying number of the "run" and another alphabet terminated
by a repetition of the identifying number. These coded
signals are recorded on the edge of the film (Fig. 13) along
the side of the photographs of the deflection on the cathode-
ray tube screen, and afford quick identification of associated
crashes. Radiotelephonic communication between Gaines-
ville and Puerto Rico is re-established and a check made to
determine whether the run has been satisfactory. Such
checks are necessary since it has been found that at Gaines-
ville the atmospherics caused by local thunder showers in
the afternoon, in many cases, cause an abundance of flashes,
thereby making it very difficult to identify those crashes
in which the observers are particularly interested.
During the earlier part of our operations, which included
the period from July 28 to 31. 1936. great difficulty was
experienced in obtaining a suitable chart for the plotting
of the bearings obtained. Chart No. 1280 of the Hydro-
graphic Office of the U. S. Navy Department (as suggested
by that office) was obtained, but the area including Florida
and Puerto Rico and the region of interest was such a small
proportion of the total area of the chart that even a mod-
erate degree of accuracy was impossible. The Florida State
Mapping Project computed and constructed a great circle
(gnometric) chart which materially facilitates the accurate
plotting of bearings from both Gainesville and Puerto Rico.
Scope of Bulletin
At this time observations can only be reported pertaining
to the hurricane passing over Florida during the last part
of July. 1936. The results were obtained by securing com-
posite photographs every fifteen minutes during the hours
from 6 PM to 8 AM. The local static (luring the time
from 11 AM to 6 PM was generally too great for reliable
observations. During the time from 8 AM to 11 AM, syn-
chronous runs were made in collaboration with Puerto Rico.
but equipment and technique for these runs had not been
sufficiently developed and the data secured were of little
24 ENGINEERING EXPERIMENT STATION BULLETIN
Some of the composite photographs obtained are repro-
duced in Figs. 11 and 12. Those chosen were taken at 7:45.
8:00 and 8:15, morning and evening, in order to correspond
as closely as possible with the times of observations of
Weather Bureau stations on the probable positions of the
Analysis of Composite Photographs
The general procedure used in analyzing the photographs
was to determine whether any ray could be found in the
approximate direction in which the hurricane was known
to be. In some cases, apparently such a ray could be readily
discerned because of its predominance, as, for example, on
July 30 at 8 I'M (Fig. 12): yet predominance alone is not
the most important criteria.
As an example of an analysis of composite photographs,
the 8 PM observation on July 28 (Fig. 11) will be con-
sidered. It would not be a very broad assumption to state
that at this time the storm was probably in that sector from
Gainesville between 135 and 180 degrees measured from
true North. The only ray showing some prominence in
this general direction occurs at an angle of 158 degrees.
However. correct azimuths of the storm can not be obtained
from a single reading, so that it becomes necessary to study
several photographs taken before and after the time con-
An example of another type of analysis, as employed in
eliminating rays which have no connection with the storm
under observation, is illustrated by the following data ob-
tained by a study of the composites taken at 5:00 PM July
29, until 9:30 AM, July 30. These exposures are shown
in Figs. 18 and 19 and are given in tabulated form in
On July 29, we knew that the storm had been at approx-
imately latitude 26 degrees north and longitude 82 degrees
west at 8:00 AM of that day. It was supposed to be mov-
ing in a general northwesterly direction at an estimated
speed of ten to twelve miles per hour, which should place
it at 5:00 PM (nine hours later), somewhere in the area
bounded by latitudes 27 to 28 degrees north and longitudes
THE LOCATING OF TROPICAL STORMS 25
Data Taken from Film Composites at Gainesville, 5:00 PM.
July 29 to 9:30 AM, July 30, 1936
July 29. 1936
5:00 PM Spray-h between 209 and 218 degrees. Open in center.
5:30 PM Spray-h between 211 anl 215 degrees.
6:00 PM Spray-h between 212 and 217 degrees indeterminate and converging to
7:00 PM Solid spray-h between 205 and 214 degrees. Heaviest line at 213 degrees.
8:00 PM Spray-h from 210 to 219 degrees. Heaviest at 219 degrees.
8:30 PM Intensity of spray-h diminishes and becomes narrower, about 215
8:45 PM Same as 9:30 PM.
9:00 PM Spray-h at 220 degrees.
9:15 PM Three equal sprays at 200. 215 and 235 degrees.
9:45 PM Spray-h at 221 degrees. Fair determination.
10:15 PM Spray-h at 221 degrees. Fair determination becoming better.
11:00 PM Good determination spray-h at 223 degrees.
11:30 PM Good determination spray-h at 226 degrees.
July 30. 1936
12:30 AM Good determination spray-h at 229 degrees.
1:45 AM Iarge area covered by spray-h. Has horns at 224 and 237 degrees and
area between these angles filled with rays.
2:00 AM Spray-x first noticed at low amplitude and angle of 212 degrees.
2:45 AM Ray-h. good definition at 239 degrees. Ray-x increases in amplitude
and shifts slightly to 291 degrees.
3:15 AM Ray-x decreases in intensity and shifts to 290 degrees. Ray-y is first
evident, of low amplitude at 279 de-rees. Ray-h indeterminate.
3:30 AM Spray-h covers 232 to 243 degrees with definite peaks at the limits
of this sector. Ray-x decreases in intensity and shifts to 289 degrees
while ray-y is constant in amplitude and has shifted to 275 degrees.
3:45 AM Ray-x is not evident and ray-y same as at 3:30 AM.
4:00 AM Ray-x again evident, low amplitude at 287 degrees. Ray-y same as at
4:30 AM Ray-h moderate amplitude at 243-244 degrees with a short associated ray
at 234 degrees and another unidentified ray at 258 degrees.
5:00 AM Ray-x Indeterminate. Ray-y of high amplitude at 275 decrees.
5:30 AM Spray-h, solid spray, from 230 to 260 degrees anI' low amplitude.
An unidentified ray of high amplitude at 278 degrees. May be ray-y.
6:00 AM Ray-y decreased In amplitude and at 275 degrees.
6:15 AM Ray-y of high amplitude at same angle as 6:00 AM frame. New ray,
ray-z. of high amplitude evident at 265 degrees.
6:30 AM Spray-h from 231 to 251 degrees rather open with definite peaks at 234.
241 and 251 degrees. Ray-z and aftjacent rays lunidentifiedi form a
spray from 259 to 267 degrees.
7:15 AM Ray-z of high amplitude at 264 degrees. New rays appearing from 235
and 252 degrees and apparently are associated with storm and will be
7:30 AM Ray-h indeterminate. Ray-z still strong at 262 to 263: degrees.
8:00 AM Ray-h of moderate intensity at 244 degrees. Ray-z still evident and
of high amplitude at 262 to 265 degrees.
8:30 AM Ray-h of low amplitude at 245 degrees. Ray-z very strong at 262 de-
grees. A new ray. unWdentified. at 232 degrees.
9:00 AM Ray-h, low amplitude, now at 246 degrees. Ray-z still very strong at
9:30 AM Ray-h of low amplitude at 246 degrees and ray-z very strong at 262
'NOTE: Ray-h or spray-h refers to the rays or sprays believed to be associated
with the tropical disturbance and coming from the region of the disturbance.
Ray-x. ray-y and ray z refer to rays other than the above.
26 ENGINEERING EXPERIMENT STATION BULLETIN
83 to 84 degrees west. Referring to the data tabulated in
Fig. 20 we find that at 5:00 PM there are several lines
forming a spray open in the center, covering the sector be-
tween 209 and 218 degrees. The only other ray near this
sector of interest lies at about 230 degrees. At 5:30 PM
we find that the spray has narrowed and moved slightly
to cover from 211 to 215 degrees and at 6:00 PM the spray
is converging with limits about 212 to 217 degrees. At
7:00 PM there is an almost solid spray between 205 and 214
degrees with the heaviest ray at 213 degrees. At 8:00 PM
there is little change except that the spray limits are now
210 to 219 degrees. At 8:30 and 8:45 PM the intensity
diminishes and the spray becomes narrower with predom-
inant ray about 215 degrees. At 9:00 PM ray-h is found
at 220 degrees.
The 9:15 PM exposure indicates three rays at 200, 215
and 235 degrees. Obviously, it would hardly have been
possible for the storm to have moved through an arc of
twenty degrees during this last half hour period, which
eliminates the 235 degree ray from our present considera-
tion. It is likewise reasonable to suppose that the 200 de-
gree ray is too far behind the probable location of the storm
arc. This leaves only the ray at 215 degrees to be con-
sidered. At 9:45 PM we have a spray of fair determination
at about 220 degrees, and at 10:15 PM the spray is still
in the vicinity of 220 degrees with slightly improved defini-
tion. The 11:00 PM frame indicates a ray of good definition
at 223 degrees and at 11:30 PM the ray is at 226 degrees.
The 12:30 AM July 30 photograph discloses a ray of good
definition found at about 229 degrees, which appears to
be the ray in which we are interested and which we have
been able to follow consistently in its movements since
5:00 PM of the night before.
The 1:45 AM photograph is different in that there is a
broad spray which covers the sector between 224 and 237
degrees with most definite rays at the limits of this sector.
With the 2:00 AM observation, there is the introduction
of another ray of interest, so for the purposes of clarity
we shall refer to the ray, or rays, which we have been
Figure 18. July 29 1936 PM to July 30 1936 AM
NOTE-The white dots were added to aid in identifying ray-h.
28 ENGINEERING EXPERIMENT STATION BULLETIN
studying and which we believe to be associated with the
tropical disturbance as ray-h or spray-h depending upon
whether a narrow or broad ray is observed. Other rays
or sprays will be designated by other letters.
At 2:00 AM there is little change in ray-h from the 1:45
AM observation, but we have a new ray at 292 degrees,
which we will call ray-x. At 2:45 AM ray-h is of good
definition, but low magnitude, and lies at 239 degrees, while
ray-x has increased in magnitude and moved to 291 degrees.
The frame taken at 3:30 AM shows ray-h now changed to
spray-h covering the sector between 232 and 243 degrees
with definite rays at the limits of this sector and with few
rays between these limiting angles. At this time ray-x has
disappeared and ray-y is unchanged in amplitude, but has
moved to 275 degrees. It is rather obvious from its angle.
its change of angle per unit of time and its disappearance
that ray-x is not connected with the tropical disturbance
under observation and, likewise, that ray-y is at too great
an angle to be considered. The 4:00 AM frame shows that
ray-x has reappeared and is of low amplitude and at 277
degrees. At 4:30 AM we find a definite ray at 243-244 de-
grees, which we believe is ray-h. but we also find very short
rays at 234 and 258 degrees.
At 5:00 AM ray-x cannot be found on the film, but ray-y
is of high amplitude at 275 degrees. The 5:30 AM exposure
discloses, at low amplitude, the solid spray h between 230
and 260 degrees. At 6:00 AM ray-y is still at 275 degrees
but of decreased amplitude. At 6:15 AM we find a new ray.
which we will call ray-z, definite and of high amplitude at
265 degrees; and ray-y is again visible at high amplitude
at 275 degrees. At 6:30 AM we find a spray between 234
and 251 degrees which, apparently, is spray-h again. This
spray-h has peaks at 234, 241 and 251 degrees. Ray-z and
adjacent newly formed rays form a spray from 259 to 267
degrees. At 7:15 AM ray-z is heavy and of high amplitude
at 264 degrees. Rays which correspond to ray-h, now also
appear at 235 and 252 degrees.
At 7:30 AM ray-h is indeterminate, possibly due to short
duration of exposure, but ray-z is still of high amplitude at
Figure 19. July 30 1936 AM
NOTE The white dots were added to aid in idt-ntifyinv ray-h.
30 ENGINEERING( EXPERIMENT STATION BULLETIN
262 to 265 degrees. At 8:00 AM we find a moderate ray
at 244 degrees, probably ray-h; at 262 to 265 degrees there
is a high amplitude ray, unchanged from 7:30 AM. At
8:30 AM we find a low amplitude ray-h at 245 degrees.
9:00 AM discloses a low amplitude ray-h at 246 degrees
and ray-z is still very strong at 262 degrees. At 9:30 AM
the low amplitude ray-h is still evident at 246 degrees and
ray-z is quite heavy at 262 degrees. This is the final ex-
posure taken during this particular period July 29 to July
30. At approximately 10:00 AM the camera is changed
from automatic operation on composite recordings to syn-
chronous runs in cooperation with the University of Puerto
From the foregoing analysis, it seems probable that ray-h
is the ray most likely associated with the storm under ob-
servation, since it persisted during the 16 hour period
analyzed and moved through those angles corresponding
with the approximate storm location. It is also evident from
the sudden appearance, rapid rise and fall of intensity,
irregular changes of angle and sudden disappearance that
ray-x, ray-y and ray-z are most likely associated with thun-
der storms of local character. It is of interest to note that
if a study had been made with only the films taken at 8:00
PM, July 29, and 8:00 AM, July 30, available the ray most
probably chosen on the latter date as being associated with
the tropical disturbance would have been ray-z, the pre-
dominant ray. which was of high amplitude and good defini-
tion and which was in the general direction of the storm
area. However, from the analysis made. it seems evident
that ray-z is caused by some form of local disturbance.
If accurate data were available on thunderstorms occurring
at that time it might be possible to locate the source of this
ray. In our preliminary studies of the rays associated with
this tropical disturbance under observation, ray-z was
thought to be associated with the storm area.
Figures 20 and 21 show bearings obtained by the Gaines-
ville station in comparison with the probable positions of the
THE LOCATING OF T'OrPI(CA STORMS
storm as charted by the Jacksonville and New Orleans
Weather Bureau stations. It should be noted that the hear-
ings obtained agree very closely with the reported observa-
tions of the Jacksonville Weather Bureau station. Only one
discrepancy exists, namely, at 8:00 AM on July 30. At this
time the location of the storm as given by New Orleans
and by Jacksonville differ by about forty miles, while our
observations give a bearing falling between these two posi-
tions. This chart was obtained from composite pictures
by methods of analysis as previously outlined.
Observations Made During Tropical Disturbance of
July 28 193: to July 31 1936
= J: A._- 8- 2 -
Iot. I.fa il |In, o chla. tI)irtre.S Drlrrceth-
July28 :00o PM 25 -1"' 80'-3O' 15X 15X
July 29 8:00 AM 2'-10' 82'-20' 180 181
July 29 8:00 I'M 27 -50' 8: -58' 218 219
July 30 8:00 AM 2. -00' 85'-00' 255 244
July 30 8:00 AM 28"-:0' 8-.'--10' 241" 214
July 30 8:00 PM 29 -.15' 85 -,15' 273 273
July 30 8:00 PM 2'.!-25' 85"-35' 267 273'
July 31 8:00 AM :30 -25' 86'-10' 284' 284
While the results obtained in this paper indicate the
probable feasibility of charting storms by electrical meth-
ods, it must be realized that these observations are based
upon a single storm. Furthermore, it is improbable that
storms can be accurately followed with only one station in
operation. At least two, and preferably three or more sta-
tions, are necessary so that, by triangulation, the actual
positions of storms can be determined. It seems advantage-
ous for these stations to be located only a few hundred miles
apart. Probably, for charting the storms in and near Flor-
ida, stations at Havana, Cuba, and in Pensacola and Miami,
Location Plotting Map for Giving Geographic Positions from Gainesville, I'la. and
San Juan, Puerto Rico
Straight .ines Radiated from Aboive Points Are Great Circles.
a9 /aoi 87 8 as8 84 83 82 6' 8 I TROPICAL DI
S! I I I.Y 2 -:
S, Ietters indicate
I I I tain'ed fr ao detect.
I station with corre
I'Il. *ati., it.I s ol)bt.inC
I'tille "nire of U.S.t
SD' and i ** I' orr<
I1' llrnl ion. a obhtnine
I~A I~nilis office lr U. S.
A 8 PM July
B 8 AM July
C 8 PM July
D 8 AM July
E 8 PM July
F 8 AM July
S I GAINESVLLE \
I 5 I i
I00 i i II.
... .' 1901 1
"* -,^ .,..,I L,^ :
azimuths as oh-
or at Gainesville
(I from .ackson-
spiond to storm
(I from New Or-
THE LOCATING OF TROPICAL STORMS
Florida, can best supplement the Gainesville station. On
the other hand, the station at Puerto Rico serves to give
valuable information when the storm is some distance away.
Other stations, possibly located at Trinidad or Jamaica.
might also be very useful for following paths of storms be-
tween Puerto Rico and Florida.
With more than one station in successful operation, it
should be possible to augment the data secured from com-
posite photographs with that taken from synchronous runs
on individual crashes and then by triangulation, determine
the points of origin of the static. From data so obtained,
at least three divisions of static could probably be made:
(a) that which comes from a fixed direction and is due
to man-made apparatus, (b) that due to fast moving local
storms, and (c) that due to slow moving disturbances, such
as hurricanes. With such information available as to the
source of the individual static crashes, less will be left to
chance in determining those emanating from the hurricane.
While the data obtained in this paper indicate that the
observed rays gave, in general, an indication corresponding
to the center of the storm, it is not definitely proved that
this be true. since frequently even the positions of the
storm as given by the Weather Bureau are not accurately
enough known for forming such a conclusion.
Furthermore, it cannot, as yet, even be said that static
emanates from all tropical disturbances. It does appear,
based on the data obtained, that for this particular storm
such radiation was obtained. However, even though local
static might in some cases be quite pronounced, in almost
every case static could be found coming apparently from
the hurricane area. A careful analysis of the photographs
taken during the passage of a storm is necessary, and con-
sideration of the movement and change of amplitude of
many rays must be investigated before accuracy can be
Some ambiguity occurred because of the inherent bi-
lateral directivity of the goniometer system as used in these
experiments, since an inspection of the photographs would
not reveal whether the static crashes came from a particu-
34 ENGINEERING EXPERIMENT STA TION Bl'LLETIN
lar position or from a point diametrically opposite. Modifica-
tions to existing apparatus will be made to secure unilateral
directivity. This will insure that the true direction is indi-
cated upon the cathode-ray tube screen and that the dia-
metrically opposite deflections are suppressed.
With bilateral directivity and with only one station in
the Florida district, each ray must be separately considered
and those not associated with the disturbance under in-
vestigation eliminated by analysis. They could also be
eliminated by synchronous runs made at several points and
Other improvements on the apparatus are also being
made. These include reduction in size of apparatus, in-
creased freedom from mechanical difficulties, brighter
images, improved photographic methods and simpler tech-
nique of operation.
The photographic technique had not been established at
the time of making these observations to the point where
satisfactory synchronous records could be obtained. It is
now believed that improvements have overcome many of
the difficulties encountered at both stations and that data
may be obtained now from such synchronous operation.
In conclusion, it must be stated that, while this method
looks promising, it has not yet been developed to a point
where successful and accurate forecasting is always pos-
sible. Some storms, either because of size, distance, or
static conditions, may never be followed by this method.
On the other hand, as improvements are made, it is prob-
able that the Weather Bureau may secure data from this
apparatus which will greatly augment that received from
I. HREORT. (rF THE RADIO RH KEAHCII B OARD (Great Britain).
11. THE CATHODE BAY OscILI.ox;APII IN RADIo RESEARI. II. la look). R. A. Wats.n-
Watt. J. F. Heard and 1.. 1. Itrainbridge-Bell. H. M. Stationery office. 1933.
III. WEATHIir PIIENOMENA ANI DIRECTIONAL OBSE VATIONS OP ATMO.sPIERICS. S. W.
Dean. I'roccrdingas f thr. Instlitlut, of Radio Engiinrters, vol. 17, pp. 1155-1191,
IV. ATMosrPIIElIC IN AISTRAI.IA. H. Munro andil I. (. H. Huxley. Radio Re-
search Itoard (Australia) Report No. 5.
V. FIEII S CAUSED BY REMOTE T'IIHINDIKIISTokRMS K. E. would Electrical Enllineeringl
vol. 55. pp. 575-5I.1 J n. une 1936.
VI. TIlE LOCATION OF TROPICAL. STORMS I t MEANS OF TIlE CATIHOD.ERAY OSCILLO-
(GAPIHl. Wayne Muaon. Florida Engineerinvt So.ciety Bulletin 12. pp. 3-5.