490 
Although propagation of 10-cm waves to distances 
much beyond the optical range had been observed 
under favorable conditions nearly a decade earlier, 
it was the striking increases in range of decimeter 
and centimeter wave coastal radars in southern Eng- 
land, observed in the summers of 1940 and 1941 re- 
spectively, which led to a concentrated attack on the 
long-range aspects of the problem. About the same 
time a need arose for more accurate knowledge of both 
‘the short-range “interference” field and the long- 
range “diffraction” field for certain communication 
projects, and the radio equipment developed to meet 
this need formed a nucleus round which the later and 
more ambitious experiments grew. The reviews of van 
ious experimental and theoretical aspects of this work 
are listed in references 1 through 15. 
Continuous observations have been carried out over 
a range of optical and nonoptical paths across the Irish 
Sea on S and X bands and over a single 38-mile land 
path on S band. These are discussed below. In addi- 
tion to this work several investigations of more specific 
propagation problems have been carried out during 
the summer of 1944. 
1. Measurements on two wavelengths in S band, 
over a 70-mile sea path between a site in South Wales 
and the summit of Snowdon (3,500 ft). This optical 
path was studied to obtain data on the probability 
of missing aircraft on S-band radars under conditions 
favorable to trapping at low levels over sea. 
2. Measurements on a wavelength of about 3144 m 
over a 90-mile sea path, with heights such that the 
path length was about twice optical range, to provide 
quantitative data on the importance of refraction in 
this waveband. 
3. Raddr measurements from Llandudno, North 
Wales, with the Isle of Man and the Irish Coast as 
the main targets, on 8, X, and K bands. The object 
was to obtain practical data on the relative perform- 
ance of K band under a variety of meteorological 
conditions which were studied simultaneously with 
the radar observations by ship, balloon, and aircraft 
measurements. Some further reference to the results 
of (1) and (2) appears below; an interim report on 
(3) has been circulated.?? 
Irish Sea Measurements 
The first plant for simultaneous measurements 
within and much beyond the optical range on wave- 
lengths of about 9, 6, and 3 cm, using heights of 
about 100 and 500 ft each site, was made in the 
latter part of 1941. 
The work was planned on an inter-service basis, 
with equipment provided by Admiralty (developed 
under Admiralty contract by General Electric Com- 
pany Research Laboratories from that used in the 
early communication studies mentioned above) and 
stations provided and operated by Signals Researct 
and Development Establishment, Ministry of Supply. 
Arrangements were made for analysis of the data by 
the National Physical Laboratory, which has also 
more recently undertaken the development of moni- 
toring equipment. The collaboration of the Meteoro- 
logical Office was received at an early date, but it 
was only when the study of the subject had made 
further progress that the need for detailed low-level 
meteorological measurements was realized 5 these have 
been undertaken by the Naval Meteorological Service, 
soundings being made in ships and by means of ship- 
borne balloons. Additional arrangements have recent- 
ly been made with the Meteorological Office for regu- 
lar aircraft soundings over the path. 
Some difficulties were encountered early in 1942 
in finding sites for the stations which were accept- 
able from all points of view, and the field work done 
in that year consisted of several short-period trials 
over a rather wide variety of land and sea paths. In 
spite of many limitations, in particular as regards 
detailed meteorological data, the general conclusions 
reached in these trials? have been largely substan- 
APPENDIX 
tiated by later measurements. Table 1 gives details of 
the sites finally adopted. 
The path length from South Wales (A and B) to 
North Wales (C and D) is 57 statute miles and that 
to Scotland (E and F) is 200 statute miles. The path 
lengths in terms of geometrical optical range for the 
eight possible paths are shown in Table 2. 
In the original scheme all the paths were to be 
collinear, but this could not be realized with the 
sites finally adopted; the South Wales to Scotland 
paths differ by about 17° in bearing from the South 
Wales to North Wales paths, the bearing of the 
former being within a fraction of a degree of true 
north. A scheme for recording data over all paths 
(though necessarily not continuously) was evolved; 
each transmitter beam was aimed for half the time 
along each of the two bearings 17° apart (a 74- 
minute period was found the most satisfactory, and 
a small change in frequency (5 to 10 mc) was made 
automatically when the beams switched over. 
At each frequency the transmitted signal consisted 
of square pulses, at equal on/off ratio, with a repeti- 
tion frequency of 1000 c. The “standard” power out- 
put in the “on” period was 0.6, 0.3, and 0.15 w for 9, 
6, and 3 cm, respectively; the signal records were 
corrected for any significant departure from these 
powers. Paraboloid mirrors 48 in. in diameter were 
used for all transmitters and receivers; these were 
mounted inside the stations behind large canvas- 
overed “windows.” The increase in mirror gain with 
frequency more thai made up for the reduction in 
transmitter power, in spite of the less effective utiliza- 
tion of the mirror area. In the receivers the 1,000-c 
component of the modulation was rectified to operate 
the recording milliammeters. Provision had been made 
for monitoring the field radiated from the trans- 
mitters and the sensitivity of the receivers, in terms 
of a standard radiated field. This scheme was brought 
‘into operation as the National Physical Laboratory 
equipment became available; other less complete 
methods of monitoring the transmitters and check- 
ing the receivers had been in operation from the start. 
(Data for the 5-cm equipment are included here 
although, as will be noted, it was not used.) 
Radiotelephone communication between the North 
Wales and South Wales stations has been maintained 
satisfactorily for two periods of several months each 
using first S- and later X-band equipment, essen- 
tially the same as that used for the signal measure- 
ments, arranged for duplex operation. A meter-wave 
system (which gives more continuous service over 
long nonoptical paths) is now being installed by Ad- 
miralty Signal Establishment to link all the stations; 
it is already operating satisfactorily over the 57-mile 
path, and a relay link from North Wales to Scotland 
is being provided. 
On S band, operation on all four links across the 
57-mile path commenced in November 1943, although 
the two from Station A (high site) had been running 
since July. During the preliminary period, up to the 
beginning of 1944, in which a number of practical 
difficulties had to be overcome, the radio results were 
subject to rather more uncertainty than was the case 
in the earlier measurements where a concentration of 
experienced personnel was possible for the short peri- 
ods involved, and detailed analysis of these results has 
not yet been attempted. One S-band receiver was in 
operation in Scotland (Station F, low site, 200 miles) 
from the end of August 1943, but apart from one 
brief period during September, no signals were re- 
ceived until March 1944, just before the second S- 
band receiver (Station E, high site) was installed. 
On X band all the stations were in operation by 
July 1944, operation on the 200-mile links having 
started a,month earlier. 
After a few months of operation of all 16 links it 
was realized that the available effort would not be 
sufficient to cope adequately with the tasks of editing 
and examining the signal records. Consequently a 
rather drastic reduction of the centimeter wave pro- 
gram was agreed to for a trial period of 6 months, 
starting October 1, 1944. For this period the follow- 
ing links were operated continuously (without beam 
switching) : on S band, A to C and D (57 miles) and 
B to E (200 miles) ; on X band, A to C and B to D 
(both 57 miles). (The possibility of a link from B 
to D on S band with separate equipment was also 
envisaged.) 
It was agreed to postpone operation on 6 em, but 
at least one 3%4-m link over each of the two path 
lengths would be added; preliminary measurements 
on this longer wavelength were already being made. 
In addition, K-band equipment for at least the optical 
57%-mile path was to be installed at an early date. 
Figure 1 shows a general view of the equipment in 
oue of the stations (D). The X- and S-band receivers 
are in the center of the picture, with the mirrors and 
canyas-covered windows behind. ‘Nhe S-band signal 
generator aud monitoring equipment are on the small 
table beside the S-band receiver. The recorders are 
mounted on a temporary table (now 1eplaced by the 
central control desk), extieme right. The empty bay, 
extreme left, was designed to house the K-band equip- 
ment. The meter-wave equipment was mounted in an 
adjoining room. 
In addition to the radio measurements, some study 
was made of the behavior of a light beam over the 
5%-mile path dwing the summer of 1944 in the hope 
that this might provide useful information on the 
refraction produced (nearly) by temperature gradient 
alone. Measurable changes in elevation were some- 
times observed by means of a theodolite, but the in- 
cidence of adequate visibility was small, and little 
quantitative information was obtained. 
A detailed study of the S-band signal records and 
meteorological data obtained from February 1944 is 
being made at the National Physical Laboratory, par- 
ticularly for the 57-mile paths AD and BD.? Similar 
study of the S band and 3%4-meter data will follow. 
Figure 2 shows a plot of hourly mean signal level 
for the S-band signal over the links AD and BD for 
June 1944, with a record of some meteorological fac- 
tors—fronts, precipitation, and fog—with which com- 
parison has been made. 
It should be emphasized that analysis of the data 
obtained during this period has not yet been com- 
pleted, but the following general conclusions may be 
drawn : : 
1. There is general agreement between signal vari- 
ations for the two paths, though the short-period 
variations often differ. 
2. Signals are obtained over the 200-mile path 
only when signals over 57-mile path BD exceed about 
30 db above 1 pv. But if the latter condition is ful- 
filled the former does not always follow. 
3. There is a marked diurnal variation, when the 
general level is low or moderate, with high signal in 
the late afternoon or evening and low level in the 
early morning. 
4. There is evidence for an appreciable seasonal 
variation with high level for a greater fraction of the 
time in summer than in winter or spring. 
5. Low level occurs commonly in conditions of 
fog or low visibility (e.g., low level on 174 occasions 
out of 233 on which fog was recorded between Feb- 
ruary and June 1944). 
6. Low level is usually observed at the passage of 
fronts (e.g., on 78 occasions out of 106 on which 
fronts were recorded). 4 
%. While periods of hign level are sometimes chiar- 
acterized by large gradients of water vapor (sound- 
ings usually made for the first 200 ft, at one point near 
the center of the path), no satisfactory correlation 
has been found between 'the character of the M curve 
and the major variations in signal level for the peri- 
ods which have been studied. In general, as is com- 
mon experience for similar paths, high levels tend to 
occur in anticyclonic periods. 
The general character of the S-band signal varia- 
