In addition to the shore soundings made at the 
water’s edge, a few soundings were obtained inland, 
in an effort to determine how far in over the land the 
duet extended. Unfortunately, most of the data are 
sparse and not too reliable. A few good soundings 
were obtained about 1 mile inland, an’ example of 
which is shown in Figure 13. The data were taken dur- 
ing the day and show clearly that no low duct existed 
at that time. This slide is a composite between a 
sounding made on a 50-ft windmill and a kite sound- 
ing made nearby. The kite was flown to 600 ft and the 
M curve continued at the slope representing mixed air 
from 60 ft on up to 600 ft. No night measurements 
were made. 
Tt was possible to make a few shipboard soundings 
to leeward of the islahd, beginning at a distance of 24% 
miles and continving on out to 20 miles. A preliminary 
study of the results shows no appreciable change over 
the course and no difference between conditions to lee- 
ward and to windward of the island, indicating that 
the duct is restored very close to shore. 
Some plots of certain correlations between wind 
speed, duct thickness and J/ deficit follow. The graphs 
in many cases are composed of very few points and 
due to the short time available are based on average 
soundings which have necessarily been smoothed. 
Figures 14, 15, 16, and 17 are based on the mean tower 
soundings and mean winds for each run, these being the 
only smooth data readily available for quick analysis. 
Figure 14 shows effective M deficit plotted against 
wind speed. This portion of the curve seems sensitive 
to wind speed variation. 
Figure 15 shows the effective slope (height of min- 
imum M divided by effective M deficit) plotted against 
wind speed. Some connection between the two quan- 
tities is indicated. 
In Figure 16 the height at which M is a minimum 
is plotted against wind speed. The isopleths of effective 
M deficit have been sketched in. A few of the points 
were thrown out in drawing the isopleths. For constant 
duct height, the effective M deficit apparently first in- 
creases with increasing wind speed and then decreases. 
Unfortunately there are only two points in the low 
wind region to establish this behavior. It is quite pos- 
sible that the lines should be more nearly horizontal 
at low wind speeds and then should slope off in the 
manner shown for winds above 15 knots. 
An attempt to plot sea temperature minus air tem- 
perature against wind speed showed no correlation 
Plotting mixing ratio based on saturation at sea tem- 
perature minus mixing ratio computed from dry and 
wet bulb temperatures against wind speed also failed 
to show any correlation. 
Figure 17'is a plot of total M deficit versus wind 
speed, with isopleths of total slope, that is, the duct 
height divided by the total M deficit. Again, the exact 
pattern of the isopleths is not definitely determined. 
With the inclusion of more data in the form of 
smoothed individual soundings, this chart and the 
previous ones may prove to be more conclusive. If this 
is the case, it may then be possible to estimate the 
values of duct height and effective M deficit simply 
from single observations of air temperature, air hu- 
midity, sea témperature, and wind. Psychrometric 
observations taken at a height of from 30 to 60 ft 
above the water would provide the value of M at the 
top of the duct to +1 or 2 M units at the most. An 
observation of sea temperature leads directly to the sea 
surface value of M, and the wind speed can be obtained 
from the ship’s anemometer. Thus with the aid of the 
charts three important points on the M curve can he 
obtained, namely, the values of M at the sea surface 
and at 1 ft and the minimum value of M and its 
height. 
These preliminary results may be summarized as 
follows: 
1. A surface duct between 40 and 50 ft high with 
a slightly transitional-type layer extending above the 
duct to between 100 and 150 ft exists most of the 
time over the water in this area. 
APPENDIX 
2. The duct is destroyed over land in the daytime 
within about ¥% mile of the shore. 
8. Islands comparable in size to Antigua have little 
effect on the duct on the leeward side at a distance 
greater than 2% miles off shore, 
~ 4. The higher the wind speed the thicker the duct 
becomes and the less the effective M deficit becomes. 
5. Changes in wind speed have little effect on the 
total M deficit, which is determined essentially by 
the temperature and humidity of the air mass as a 
whole in relation to the surface water temperature. 
6. These conditions probably prevail over ocean 
areas having comparable climates, 
Preliminary Results of Radio and Radar 
Measurements? 
The main purpose of the experiment was to estab- 
lish what operational use could be made of low-lying 
ducts and to confirm observation of the effects of such 
ducts on radio and radar propagation made in various 
parts of the world. The data accumulated have been 
‘available for study only 2 weeks, and there has been 
insufficient time for a complete analysis. As a con- 
sequence only the highlights of the agreement between 
experiment and theory have been determined. 
Ducts were present all the time, and trapping on 
both X and § bands, which increased the signals to 
levels considerably above standard propagation values, 
was found to exist all the time. The general conclusion 
regarding the effect on the two bands was that on S 
band antennas as high as the experiment would allow 
gave the highest signal strengths. On X band, on 
the other hand, the lowest antenna heights which were 
available usually gave the strongest signals. 
Figure 18 is an S-band run made on March 15. It is 
a composite run containing the results of both the 
outward and the inward runs. Several of the curves 
have been omitted for clarity. The highest curve is 
for a combination of a 46-ft transmitting antenna- 
and 94-ft receiving antenna. The lowest curve is for 
the two lowest heights, 16 and 14 ft. The slopes of the 
curves are rather steep for the first 80 miles or so, 
the signal declining considerably less rapidly there- 
after. Also, the variation of the signal with height is 
shown here to be in the order in the extremes between 
25 and 30 db. This interval from 80 to 50 db shows a 
difference between the two extremes of 30 db. To trans- 
late that into a radar situation, double that difference 
to get a difference of 60 db, showing that on S band 
the higher antenna combinations would provide con- 
siderably better coverage for targets in the order of 
100 ft high and with transmitters at the height of about 
50 ft. Stated another way, the highest antenna com- 
oination would provide coverage beyond that obtain- 
able with the lowest in the order of 30 miles. 
There is as yet no reasonable explanation for the 
extremely slow decrease in signal beyond 80 miles. 
This feature is very distinctive in the S-band curves. 
For the X band, it is generally not discernible except 
on a few runs toward the extreme range portion. The 
tate of decrease of signal with range in the region. 
inside 80 miles would be exponential if there were a 
straight line on this figure. Considering it to be s0, 
averaging over a number of runs gives roughly 0.8 db 
per nautical mile. That decrease is the total amount, 
the 1/R variation not having been extracted from it. 
Attempts to do so show that the resulting curve does 
not, in a plot of this sort, fit a straight line as well 
as the original values themselves, but if the 1/R value 
is taken out of the power relation the average attenua- 
tion is then roughly between 0.5 and 0.6 db per nau- 
tical mile. In this region (beyond 80 miles), on the 
other hand, the decrease of signal with range is con- 
siderably less, being between 0.15 and 0.2 db per nau- 
tical mile. No satisfactory explanation for this be- 
havior has yet been derived. 
Figure 19 shows the X-band results for the same 
period. Antenna heights of 16-ft transmitting and 
‘By M. Katzin, U. S. Naval Research Laboratory. 
489 
6-ft receiving produced the highest curve, the lowest 
curve being obtained on a 46-ft to 94-ft combination. 
Note that succeedingly higher antenna combinations 
produced successively lower signal strengths. There is 
some variation, but when the curves gre smoothed 
to a straight line the attenuation is on the order of 
0.33 to 0.5 db per nautical mile. Removing 1/R re- 
duces the attenuation to roughly 0.2 db per nautical 
mile. There is no sharp bend in the curve at about 80 
miles, as was the case on the S band. The lowest (16- 
ft to 6-ft) antenna combination showed more than 
35 db greater signal strength than the highest (46-ft 
to 94-ft) combination. Considering again the radar 
case, it is found that the higher antenna provides rela- 
tively poor coverage compared to the lower. In terms 
of range for a given signal threshold, the difference in 
favor of the lower antenna is about 80 miles. 
Figure 20 shows an X-band curve obtained during 
April 10 and 11, when a transmitting antenna height 
of 8 ft was available. Received signal powers for 6-, 
14-, 24-, 54-, and 94-ft receiving antennas are shown. 
The curves are somewhat scrambled, but the general 
result is that the lowest antenna again produces the 
greatest signal, with increasing antenna height pro- 
ducing progressively smaller signals. This was not the 
case without exception, as can be seen in Figure 4, 
where the 6- and 14-ft antennas exhibit comparable 
behavior. In that case the maximum range was ob- 
tained on the 14-ft antenna. The average slope in 
Figure 20 is somewhat less than that shown in Fig- 
ure 19. Exact averages of all the runs have not yet 
been worked up. 
Figure 21 shows a plot of received signal versus 
range, made on a 3-cm radar, using a PC boat as a 
target. The highest curve was obtained with a 6-ft 
antenna height, using a 48-in. dish to obtain greater 
gain and range. The other run with 6-ft antenna was 
made using the regular 29-in. dish. There is a consider- 
able spread in the values of received signal due to the 
difficulty of measurement. However, the significant 
thing is that the maximum ranges obtained are in 
accord with the indications given by the one-way trans- 
mission results. Striking an average slope shows the 
decrease of signal with range to be about 1.0 db for 
each 1.5 nautical miles. 
The important conclusions can be summarized as 
follows: 
1. The surface duct is very persistent. 
2. The duct is very effective in extending the ranges 
obtainable on both S and X bands, for either one-way 
or two-way transmission. 
3. On S band, the highest combination of trans- 
mitting and receiving antennas produces the strong- 
est signal and the greatest range. 
4. On X band, the lowest combination of transmit- 
ting and receiving antennas produces the strongest 
signal and the greatest range. 
5. Surface ducts in the trade wind regions can be 
used for communication purposes to a conservative 
range of 100 miles. Greater ranges are probable but 
will require further investigation. 
6. Rain in the form of squalls does not appreciably 
affect the received signal. 
ENGLAND 
BRITISH TRANSMISSION 
EXPERIMENTS: 
Introduction 
HE BROAD OBJECT of the studies carried out in 
Great Britain during the past few years has been 
to establish the characteristic facts of the propagation 
of centimeter waves (more recently of meter waves 
also) and especially to determine the relationship be- 
tween radio performance and meteorological condi- 
tions in the lower atmosphere, with forecasting as the 
ultimate aim. 
"By E. C. S. Megaw, Ultra Short Wave Panel, Ministry of 
Supply, England. 
