UNIVERSITY OF FLORIDA PUBLICATION
Engineering Experiment Station
Bulletin No. 6
ON STATIC EMANATING FROM
SIX TROPICAL STORMS AND ITS USE IN
LOCATING THE POSITION OF THE DISTURBANCE
S. P. SASHOFF
Dean, College of Engineering
Director, Engineering Experiment Station
Reprinted from Proceedings of the I.R.E.
Vol. 27. No. 11, November, 1939
UNIVERSITY OF FLORIDA
Static Emanating from Six Tropical Storms
and Its Use in Locating the Position
of the Disturbance*
Stephan P. Sashofft, NONMEMBER, I.R.E., AND
Joseph Weilt, MEXMlER, I.R.E.
Summary-This paper describes the technique of locating the
apparent source of atmospherics employing several stations located
at a considerable distance from each other and equipped with cathode-
ray-tube direction finders. It discusses the data obtained on six tropi-
cal storms and compares the results with the estimate of the position
of the storm center as reported on the daily bulletins of the United
States Weather Bureau. It presents some evidence of close connection
between static and meteorological conditions. It suggests that further
studies should be made along the same or similar lines to establish
the feasibility of this method for locating the center of tropical storms
by means of their associated static.
RD EPORTS of investigations on the nature of
atmospherics have appeared in print from
time to time. These investigations have dealt
with static emanating from lightning and thunder-
storms. While in most of the cases the observers have
aimed at measuring the intensity, direction, and the
wave pattern of the received static other investiga-
tors have attempted to develop some means of
locating the position of the source of the arriving
Fifteen years ago \atson Watt, Appleton, and
Herd' found that atmospherics recorded on the Con-
tinent along with European time signals were coin-
cident with the time of arrival of static recorded in
America along with the same time signals. This led
to the development of a method of locating the source
of static by triangulation, using a cathode-ray-tube
radiogoniometer or direction finder.
Decimal classification: R114. Original manuscript received
by the Institute, June 16, 1939. Presented, joint meeting,
.R.S.I.-I.R.E., Washington, D. C., April 29, 1938.
t University of Florida, Gaincsville, Florida.
Appleton, Watson Watt, and I lerd "On the nature of
atmospherics," Proc. Royal Soc.. series A, vol. 3, pp. 654-677;
The University of Florida became interested in the
study of atmospherics about four years ago. After a
series of conferences with officials of the United States
Weather Bureau, the National Bureau of Standards
and other interested parties, it was decided to at-
tempt such a study in order to determine how close
a relationship, if any, exists between static and
meteorological conditions. It was agreed that any
new means for supplying information relative to the
Fig. 1-Perpendicular rectangular loops used with
cathode-ray-tube direction finder.
position and movements of tropical storms should be
investigated, since if found reliable they would be of
great value to the people of Florida and to those of
the Caribbean countries because of the frequency of
occurrence and the great loss of life and property
which often follows in the wake of such storms. A
systematic study of the nature and origin of atmos-
pherics was undertaken, therefore, by the Univer-
sity of Florida and the University of Puerto Rico, Rio
Piedras, to determine the feasibility of locating the
center of tropical storms by means of their associated
static. The necessary funds for this work were pro-
vided by the Works Progress Administration.
Through co-operating agencies, equipment built by
the National Research Laboratories of Great Britain,
Slough, for the United States Navy was made avail-
able to the two schools. This equipment was re-
modeled and brought up to date and tests were begun
in September, 1935. A report on the earlier observa-
tions has already been published.2
TRIANGULATION FOR THE CENTER OF TIHE STORM UY
MEANS OF STATIC ASSOCIATED WITH IT
1. Equipment and Method
This paper describes the work on the oscillographic
examination of the direction of arrival of atmospher-
ics, the results of triangulation for the center of the
static source, and the comparison of the position of
the apparent source of the incoming static and that
of centers of storms determined by other means. This
series of tests began in August and continued through
October, 1937. The method used in making the ob-
servations is in general that of Watson Watt and his
Two perpendicular rectangular loops are located
close to each other; one in the true East-West and
the other in the true North-South directions as
shown in Fig. 1. The outputs of these loops are fed
into separate tuning networks. From the tuning net-
works the incoming signal passes into two amplifiers,
one connected to each tuning unit. These amplifiers
are so designed and adjusted that signals of equal
magnitude and in phase will appear at the output
terminals still equal and cophasal. The output of the
East-West amplifier is then fed to the vertical plates
of a cathode-ray tube while the North-South ampli-
fier supplies the horizontal plates. The corresponding
points of the compass are marked at the end of the
tube as E, IV, N, and S.
With this arrangement a static pulse arriving at
the loops is passed through the tuners and the ampli-
fiers and then appears on the screen of the tube in
University of Florida Engineering Experiment Station Bulle-
tin No. 3, October, 1936.
the form of a bright line. The orientation of this line
with respect to the EWNS points at the end of the
tube is that of the actual direction of arrival of static
with respect to the loops. These lines, to be referred
to from now on as static crashes, can be observed
visually. Because of their short duration, however,
visual observations are difficult. A camera is incor-
porated in the equipment, therefore, the crashes
being photographed and a study of them made from
the developed film.
Fig. 2-Composite photograph. Exposure,
30 seconds at f/11.
The camera used in these studies was especially
designed for this work by WV. Mason and others. It
uses 16-millimeter motion-picture film. The film may
be kept stationary or moved at the rate of one inch
per second. If the film is kept stationary while static
is received a photograph is obtained which contains
a large number of crashes. This type of photograph,
Fig. 2, has been called a "composite." Photographs
of static taken on moving film on the other hand
have been designated as "synchronous." The name
"synchronous" was derived from the fact that such
photographs are usually made synchronously by all
stations of the network. In order to synchronize the
films at the various recording stations, radio signals
in the form of letters of the alphabet in the continen-
tal code, as shown in Fig. 3, are sent from the master
station at Gainesville, Florida. The radio signals are
picked up at the recording stations by short-wave
receivers, the outputs of which are fed into neon glow
lamps. The flashes of the lamp at each recording sta-
tion correspond to the dots and dashes of the radio
signal and are recorded on the film side by side with
the static crashes. In addition to the synchronizing
impulses and the record of the static the film record
also contains the number of the test run, the time of
day, and the date.
2. Analysis of Data
A study of the data thus obtained is made in the
following manner. The film record of a given run is
projected on a screen on which a compass rose and
rose and scale. Note direction of static crashes and synchroniz-
ing signal in continental code.
which also is the zero line, coincides with the center
S -- --W'
Fig. 3-Section of moving-filn record superimposed on a compass
rose and scale. Note direction of static crashes and synchroniz-
ing signal in continental code.
a scale having 170 divisions are drawn. T'he projector
and the film are adjusted so that the point at which
the line of the static crash cuts the eastes- t line,
which also is the zero line, coincides with the center
of the compass rose; the angle of the static crash is
measured and the end of the letter of the synchroniz-
ing signal just above the crash noted on the scale.
The readings obtained in the above manner are
recorded on a form known as "the static record
sheet." Such sheets are made for the individual runs,
one for each of the recording stations. While analysis
of the film records described in this paper were made
separately at Gainesville and at Rio Piedras, the
method may be used to read the film records at each
station and then transmit the data by radio to a cen-
tral point where triangulation may be made. In order
to insure identical projection of the films and to com-
pensate for any differences in film speeds, series of
equally spaced dots are sent with each run. The per-
son making the analysis is instructed always to pro-
ject the film in such a manner that 5 of these dots
cover 100 divisions of the scale.
The 2 recording stations already mentioned began
making observations at the beginning of the 1937
hurricane season. Before the end of the period two
additional stations, one at Miami and another at
Pensacola, Florida. were put in operation.
The film records of the three Florida stations were
developed in Gainesville, while G. W. Kenrick and
his staff at the University of Puerto Rico developed
their own films. Records were exchanged by mail.
When analysis of data first began it was our inten-
tion to look only for "synchronous crashes"; i.e.,
crashes which appear under corresponding points of
the synchronizing signal. It became evident early in
the work, however, that the number of such crashes
coming from the direction of the storm is very small in
comparison with the total number of clashes re-
corded. At first it was tllouglit this may be due to
some error possibly introduced by the unequal time
intervals required for the synchronizing signal to
reach each one of the recording stations, as well as
by the unequal time delay in the various receiving
equipment. A thorough check for a possible correc-
tion factor, which should be a constant for each sta-
tion, failed to produce any satisfactory results.
A further study of the data taken during days
when a tropical storm was known to exist in a certain
area, however, showed a considerable amount of
static arrived at each station from the direction of
A new tabulation of the static crashes was then
made. In this tabulation the crashes for each run
were segregated according to the direction of their
arrival in groups within ten degree angles. The list-
ing of the data was made in 18 groups, the angles
being measured from 355 to 175. Each group was
designated by the angular position corresponding to
the center of the group angle. The synchronizing
signals were disregarded.
The graphs thus obtained were then plotted on
"static continuation sheets", 9 graphs to a sheet. A
study of the film records and the "static continua-
tion sheets" leads to the following conclusions:
1. There is a random distribution of static crashes
especially during the afternoon and early evening
runs which suggests that most of the disturbances
during this period are of local origin. (Fig. 4.)
2. Considerable amount of static arrives from the
direction in which a storm is known to exist.
Fig. 4-Random distribution of static crashes. Storm at 120
c degrees s fror i ainesville, Florida.
.11 I' '-, '-p
i _i- I -T
tion varies, such static seems to persist on succes-
sive runs. (Fig. 5.)
4. The direction of the persisting static changes in
close relationship to the shifting of position by the
5. Static from directions like the above seems to dis-
appear with the disappearance of the storm.
appear with tie disappearance of the storm.
6. Many of the crashes in a given direction occur in
groups of twos and threes. This supports the find-
- '4- 1Y
Fig. 5-Storm static persists on successive runs. Storm at 120
degrees from Gainesville, Florida.
ings of Appleton and Chapman,3 who concluded
that if atmospherics originated in thundercloud
SAppleton and Chapman, "On the nature of atmospherics,"
Proc. Royal Soc., series A, vol. 158, pp. 1-22; January, (1937).
discharges they may be expected to have the tend-
ency to occur in groups.
7. The intervals between the individual crashes oc-
curring in a group were found to be of the order of
8. The best records were obtained between the hours
of midnight and 12:00 noon.
3. Paths of Six Tropical Storms
Having outlined the general method in securing
the data and having listed the conclusions drawn
from the compiled results, we shall view the results
obtained on 6 tropical storms which occurred during
the 1937 hurricane season, daily reports for the posi-
tion of which were supplied by the United States
Weather Bureau. At this time it should be pointed
out that observations on these storms were made for
periods of several days. At first runs were taken on
these storms every 2 hours on the odd hour, but later
because of interference from local static, the schedule
was changed and runs were taken every hour on the
hour from 7:00 r.M. until 9:00 ..M. of the next day.
The observations were made with the equipment
tuned to 10 kilocycles, this frequency having been
found by previous investigators4 to be the most
common to this type of disturbance.
The first storm was detected by the United States
Weather Bureau station located at San Juan, Puerto
Rico, on August 24. The storm had its origin several
hundred miles southeast of the island. The four sta-
tions of the network were put on a 24-hour schedule,
making observations every other hour on the odd
hour. The crashes received at the Gainesville station
definitely indicated a continuity of static from the
direction of the storm.
As the storm approached the Florida coast, the
triangulation from Puerto Rico was of little value
because the storm was nearly in line with the 2 sta-
As the storm passed the Florida shore line, squally
Harald Norinder, "Cathode-ray oscillographic investigations
on atmospherics," PRoc. I.R.E., vol. 24, pp. 287-304; February,
weather was felt throughout the state. This caused a
high level of local static to persist, making analysis
difficult, but continuity of static was noted, and the
direction of the storm followed by the Gainesville
station until it dissipated itself near Lake City,
Florida, on August 29. The path of the apparent
source of static and that of the center of the storm
as reported by the United States Weather Bureau
are shown in Fig. 6.
A separate set of data was sent to the Puerto Rican
station for independent analysis on this storm, and
I ,.- 1H -T 0.E."
LT-0 AT NT ALE
C, -', -iI* .&"F
Fig. 6-Path of Atlantic gale, August 25-30, 1937.
the results obtained there were markedly consistent
with results obtained at Gainesville.
The second storm of the season occurred from
September 9 to 13. It was brought into observation
by the static-recording station a few hours after the
United States Weather Bureau had approximated its
position. This storm was of full hurricane force near
the center, accompanied by heavy rains and heavy
seas in the general area. The Gainesville observations
as early as 9:07 A.M. on September 9 definitely indi-
cated static emission from the general storm area.
The storm was followed and the results plotted as
shown in Fig. 7.
The third storm persisted from September 14
through the 15th. It was located about 360 miles
northeast of St. Martin, Leeward Islands, and was
attended by winds of full gale to hurricane force.
This disturbance did not cover a wide area and had
very little if any movement. It dissipated itself in
about 36 hours. The Gainesville and Puerto Rican
stations definitely indicated emanation of static
from the general storm area. (Fig. 8.)
The fourth storm was an Atlantic hurricane lasting
from September 20 to September 25. This storm
traversed a distance of about 2500 miles which made
W' S *. mt A LANIIICHUMKAC
9 1 37
sal V 0 ?
.. 10. 1 AM '
Fig. 7-Path of Atlantic hurricane, September 9-13, 1937.
it of considerable interest. Apparently the equipment
would record static emanating from a source several
thousand miles distant. It was observed that while
the addition of another thousand miles from the
source of static to the Gainesville and Puerto Rican
STROPIC L DISTURB KE
"S r. '1 S 4-l 1.I-
S ITATIC PATH
Fig. 8-Position of tropical disturbance, September 14-15, 1937.
stations had little effect on the number of static
crashes received, the amplitude of the crashes, as
would be expected, showed a noticeable decrease. In
general, for this as well as the other storms under
observation, the static source corresponded favorably
with the storm positions as plotted by the United
States Weather Bureau. (Fig. 9.)
The fifth storm occurring from September 27 to
28 was an Atlantic gale, located in such a position
that good triangulation was obtained from all the
stations. About the only significance that could be
placed on this disturbance is that static apparently
emanates from squalls as well as hurricanes. Fig. 10
shows the course of this storm.
'I he sixth and last storm of the 1937 season lasted
S. urt wn
I. i fa S I
*1 4 ATANTIC *I fR 5
Fig. 9-Path of Atlantic hurricane, September 20-25. 1937.
from October 1 to October 3. The results were quite
consistent with those obtained by the United States
Weather Bureau plotting. (Fig. 11.)
A study of 6 maps showing the apparent position
of the static source and the corresponding position
of the center of the storm as indicated by the Weather
Bureau shows that the two rarely coincide. Several
explanations for the discrepancy may be offered.
This may be due to one or all of the following:
a. Errors due to location of the equipment.
b. Errors due to improper location of the plates of the
c. Errors due to improper location of the film in the
d. Errors in the reported position of the storm.
Another explanation suggested by some investi-
gators and partly verified by Florida observers is
I 0.'n r . *J[ 1
#- I A' A- AT I. AI.E
that static does not apparently emanate from the
center of the storm, but rather from its periphery.
If this be the case, it remains for further studies to
show just how far from the center of a storm static
does emanate. When and if this is determined, there
I ...... An T 6A .
OCT-) 1 --3. 19.7
.... - STATIC _A H
T(* L I 0 I.
Fig. 10-Path of tropical storm in the Gulf of Mexico,
October 1-3, 1937.
still remains to be found the exact part of the
periphery from which the emission occurs. If these
factors can be determined, perhaps a fairly accurate
approximation of the storm center may be made.
The authors propose to continue the present in-
vestigation. Future tests are to include observation of
the incoming static and triangulation for the position
of the source without a knowledge as to the exact
location of a storm but working only with informa-
tion that a tropical storm does exist.
Further extension of the work will include a study
of the wave shapes of atmospherics along the lines
followed by other workers in this field. It is hoped
that perhaps hurricane static may have some char-
acteristics of its own to distinguish it from static
caused by lightning and thundershowers. If such
"fingerprinting" of hurricanes is found to be possible.
then a more accurate and rapid location of the storm
center by this means may be feasible when no other
means of determining the position of the disturbance
We take pleasure in expressing our gratitude to the
staffs of the various recording stations in Florida and
to Dr. G. W. Kenrick and his coworkers at the Uni-
versity of Puerto Rico, Rio Piedras, for their untiring
co-operation in securing the necessary data; to the
Works Progress Administration for making available
the necessary funds for carrying on the work; to the
United States Navy Department for making avail-
able to us the original Watson Watt equipment; and
to all others who have helped with the work through
suggestions and helpful criticism.