and the preceding two speaks for itself. From the pre- 
liminary analysis no consistent correlation has been 
found between the behavior of the low- and high- 
frequency transmission. 
The variations on the two paths are generally in 
good agreement although changes in signal type rare- 
ly occurred exactly simultaneously; the changes on 
the short path are always less in magnitude than on 
the other, as would be expected. 
As far as can be determined from the available 
data the K-band signal correlates quite well in gen- 
eral with that on S and X bands. Only high signal 
can be observed, of course, with the present equip- 
ment. 
Retation or Rapro Resutts ro Moprrrep 
Inpex Curves 
Detailed conclusions must await the full analysis 
of the data. At present certain qualitative conclu- 
sions can be drawn: 
1, When the surface modified index inversions are 
present, the microwave signal level is high on the 
average, and usually the signal has roller-type fades. 
_ 2. When the M curve is substandard the signal is 
low and scintillating. The M curves which are stand- 
ard all the way down to the surface of the water 
appear to be very rare, even when the air is colder 
than the water. The previous results on the short 
path had tended to discount the importance of the 
low M inversions which exist over water most of the 
time, especially with cold air flowing out from the 
land. The increased sensitivity of the present setup 
to variations in the M curve, the additional path 
length, and finally the inclusion of the X-band trans- 
mission on the circuit have shown definitely that such 
low M inversions are far from negligible but will 
affect S-band communications (or one-way trans- 
mission) and both one-way and radar:transmission on 
X band. The signal occurring with these M inver- 
sions less than 20 ft high is usually the high, steady 
type. It is generally not quite so high in average level 
as that characterized by roller fades found with larger 
M inversions. 
3. The high, steady signal occurring with very 
low M inversions reveals the only clear-cut cases of 
frequency diversity between S and X bands. In this 
case a variety of combinations has been found: nearly 
standard signal on S band with X-band signal from 
10 to 30 db above standard; S band 1@ to 15 db above 
standard with X band 30: or so db above standard; 
and finally S band about 20 or more db above stand- 
ard and steady while X band changes to the first sig- 
nal type: high with fades. 
4. One of the most interesting features of the trans- 
mission is the fact that at any given location, for a 
fixed frequency, the increase in field strength is lim- 
ited; that is, no matter how much the M inversion 
increases in height or in strength beyond a certain 
value (which is as yet unspecified) the average value 
of the signal strength does not continue to increase 
but rather remains the same within about 10 db. This 
“saturation” level is of the order of the free space 
value. (Maximum level goes up. to 12 to 15 db above 
free space but only infrequently.) Consequently, the 
level reached on a given path appears to be indepen- 
dent of the receiver height (within the height range 
covered in these measurements), the height-gain effect 
which exists under standard conditions being essen- 
‘tially eliminated when shore trapping takes place. 
Under some conditions, especially when the signal is 
high with fading but has not yet reached the satura- 
tion level, the lower of the two receivers has been ob- 
served to receive higher signal than the higher one. 
' With stronger signal the values on the two become 
nearly identical, as has been stated. 
These results agree with unpublished calculations 
made for several values of duct height and M deficit 
for S band, of the first transmission mode alone, 
which indicate that the height-gain effect should dis- 
appear and the signal approach a certain saturation 
APPENDIX 
level. Thereafter, calculations show, the contribution 
of the first mode decreases, but the observations sug- 
gest that perhaps the other modes continue to cause 
the average level to reach approximately the same 
value, as duct height and M deficit continue to in- 
crease, 
Tt has been found that, with an M inversion over 
only a portion of the path and a standard curve on at 
least a small part of it, the signal type may be high 
with roller fades and the average level high, so that 
tle record is indistinguishable from that which oc- 
curs with more uniform conditions 
Radar Transmission 
From Race Point, targets were available over water 
at ranges of 20 to several hundred miles along the 
coast of Massachusetts and Maine, plus some addi- 
tional targets inland and whatever shipping was in 
the vicinity. Of the coastal targets four were chosen 
for regular .observation. These were fairly isolated 
fixed targets, the echoes from which appeared to be 
telatively steady in several days’ observations, at 
ranges of 22, 41, 65, and 73 statute miles. Absolute 
power measurements of the returns of each of these 
targets (whenever visible) were made hourly by com- 
parison with a signal generator. Hach measurement 
represents the maximum value of the signal during 
a period of 1 to 3 minutes. This differs rather essen- 
tially from the hourly averages of the one-way data. 
In addition to signal strength measurements, hour- 
ly observations were also made of the maximum ranges 
obtained on surface targets, and plan position indt- 
cator [PPI] photographs were made which reveal 
at a glance many interesting features of the radar 
coverage which are hard to describe briefly in words. 
The maximum sweep length available on the PPI 
was 140 miles for the S-band set and 115 miles at X 
band. Additional range was available on the delayed 
A-scope sweeps, so that the maximum was 180 miles 
for most of the period of observations. This was ex- 
tended to 280 miles for the last week of the test. 
In addition a portable K-band radar set was set up 
near Race Point Light only 1% ft above mean sea 
leyel and regular observations of range were made and 
shipping tracked. 
TarGET SiaNaL STRENGTHA 
The strength of the echoes from the four targets, 
including the nearest one which is ordinarily visible 
both optically and by radar, varied from below mini- 
mum detectable to at least 60 db above for the two 
nearer targets and about 35 db above for the two 
more distant targets, at both frequencies. In general 
the values of the signal strength were higher for the 
nearer targets. but there were some interesting cases 
when the more distant targets were visible while the 
nearer ones were either not seen or were very weak. 
This may occur at times when the M curve varies 
markedly with direction, as happens occasionally when 
the air trajectory is S or SW or at times of skip dis- 
tance. 
Maximum Rances 
Large variations in the maximum ranges have also 
been observed at both frequencies, with the upper 
limit apparently being set only by the length of the 
sweep: 280 miles on S band and 200 miles on X band. 
(Note that these radar sets were far from the high- 
power class.) Lack of fixed targets at ranges between 
10 and 25 miles made it impossible to follow in detail 
the way in which substandard conditions reduced de- 
tection range, but there was no question as to the 
general trend toward reduction of range. The maxi- 
mum range of the high sited K-band receiver [HRK] 
from its location at the Race Point Light was 46 miles 
on a land target and about 30 miles on shipping. 
It should be borne in mind that our project deals 
with propagation near and roughly parallel to the 
coast line. Thus these results are not necessarily ap- 
plicable to operations perpendicular to the coast with 
off-shore winds, where the surface M inversions be- 
483 
come “washed out.” 
Sratistics 
The radar observations include about 1,200 hours 
of operation, Of these, overall, the X-band ranges were 
better than “normal” 59 per cent of the time (normal 
= 29 miles*) and the S-band ranges 48 per cent of 
the time. At both frequencies ranges were below nor- 
mal 20 per cent of the time. The variations from weck 
to week were great, the maximum values being 95 per 
cent above normal on X and 75 per cent above normal 
on §, with about 45 per cent below normal as the 
lowest value at both frequencies, 
CorrELATIONS WITH ONE-Way Resutts 
A visual comparison of the radar and one-way data 
suggests fairly good agreement in general between 
the two. To get a more quantitative evaluation of this 
agreement. however, correlation diagrams have been 
drawn. 
Figure 7 shows such a diagram relating the signal 
strength of the target at Eastern Point as observed 
on the X-band system with the signal strength of the 
high X-band receiver on the one-way path. As in the 
previous diagrams, a week has been chosen as the time 
interval and hourly values are plotted. In this we 
neglect the difference between the single observation 
of the radar and the average of an hour’s continuous 
record in the other case. Note also that all radar 
measurements which give values equal to or below 
the minimum detectable level are plotted at the min- 
imum detectable level; thus if a more sensitive re- 
ceiver had beer used, many of these points would have 
fallen lower in the diagram. The diagram reveals the 
nature of the relation: the one-way signal strength 
must rise considerably above the standard value before 
the target becomes visible. Thereafter, small changes 
in the one-way signal correspond to much larger 
changes in the radar echo. As a matter of interest, 
which may or may not be significant, the values at 
times fall close to the square law, as they should if the 
target-reflecting properties remain constant as the 
tmospheric conditions change. 
Figure 8 shows the relation between maximum radar 
Tanges on surface targets and the one-way transmis- 
sion results, In this case the effects of both substandard 
and better than standard conditions are noticeable. 
When the one-way signal strength is below standard, 
the radar ranges are mainly less than normal; excep- 
tions occur in cases of strong directional effects and 
S-shaped M curves. As the one-way signal strength 
rises above the standard level no appreciable increase 
in radar range occurs at first. Only when the one-way 
signal strength has become fairly high do the radar 
Tanges begin to increase. Then the entire gamut of 
long radar ranges, from about 40 to 280 miles, takes 
place while the one-way signal strength changes only 
slightly. This is another manifestation of the satura- 
tion of the signal at a high value. 
Summary 
Two major conclusions may be drawn from this 
preliminary survey: 
1. Standard signal is the exception rather than the 
tule for microwave radiation on this over-water path 
during the summer and fall. With the high M inver- 
sions, Which occur with warm, dry air over water, 
signal strengths 30 to 45 db above the standard occur 
about equally often on both, the upper limit being 
approximately the free space value, and radar ranges 
on surface targets are extended to five to ten times 
their normal values. On the other hand, with the low 
M inversions (less than 20 ft, say) which occur with ~ 
air colder than the water, X band is affected more than 
S. Both may experience increases in signal level of 10 
to 30 db above the standard, but the X-band signal 
*Unfortunately, in the radar case it is impossible to estab- 
lish a precise definition of a ‘‘standard” range analogous to 
standard signal in one-way transmission unless detailed in- 
formation is available on the radar target. In this case we 
have attempted to determine the detection range on the low 
hills available as coast line targets, at times when the M 
curve is standard or very nearly so. 
