310 RADIO WAVE PROPAGATION EXPERIMENTS 
from Deer Island in Boston Harbor showed a 15-db 
variation with tide. Their range is 10,000 yd, and the 
targets are about 60 ft high. Under these conditions 
the targets subtend more than two lobes of the inter- 
ference pattern'at S band. Since the effect of the tidal 
change is to move this lobe structure up and down by 
10 ft, it is difficult to believe that a change as large as 
15 db could be thus produced. However, one can break 
up the returned signal into a number of separate sig- 
nals differing as to whether they suffer two reflections 
on the surface of the sea or one reflection or go directly 
to and from the target. While the amplitude of each 
of these signals does not vary much with tide, their 
relative phases do, and the total signal can still change 
considerably in amplitude because of the interference. 
A similar set of measurements was made on a corner 
reflector mounted on a small island in Boston Harbor. 
Here the corner reflector although only 6,000 yd away 
acts essentially as a point target. The agreement with 
the theory for a point target is quite good. 
Our results emphasize the extreme caution that must 
be employed in the use of standard targets to monitor 
radar performance. They are just the type of targets 
which are normally chosen in the field, and obviously 
their variations with the tide make them entirely un- 
suitable for the purpose. It may be possible to find tar- 
gets whose echoes are sufficiently steady so that they 
can be used for monitoring. However, they cannot be 
found without the use of such test equipment as would 
obviate the need for standard targets. 
THE FREQUENCY DEPENDENCE 
OF SEA ECHO? 
As the power and frequency of radar sets continue 
to increase, and the size of the target to be detected 
decreases, the presence of sea echo becomes of ever 
greater operational significance. It acts as a built-in 
jammer, blanketing and obscuring the desired signals. 
Despite this growing practical importance the basic 
phenomena of sea echo have not yet been established. 
Certainly, the fundamental mechanism responsible 
for the signal is not yet known. Various conflicting 
theories have been proposed. It has been suggested 
that scattering from drops of spray is the cause of 
the echo. Another hypothesis is that of reflection or. 
diffraction from the large surfaces of the waves them- 
selves. Still other theories have been advanced at one 
time or another. 
Whatever the size of the scatterers, the power re- 
ceived at the radar can be described by a common 
formula. Consider some particular scatterer, say the 
jth one. Then the returned signal from this particular 
target is 
eed G? 2 
" (4m)3 RA 
bBy H. Goldstein, Radiation Laboratory, MIT. 
oj, 
where a; is the radar cross section of the jth scatterer, 
P, the power transmitted, G the gain, X the wave- 
length, and & the range. o; differs from the customary 
cros$ section in that it incorporates the propagation 
factor and hence may depend on the height of target 
and the glancing angle of the incident ray. In consid- 
ering the time average of the total power received by 
the radar we can take the scattering to be incoherent. 
Hence the average radar signal is the sum of P,, over 
all the j scatterers lying within the area illuminated 
by: the beam width and pulse length: 
La S\ 03 
pie es Se 
(41)3 R44 
It is assumed that the illuminated area is sufficiently 
large that the sum contains many scatterers and is 
proportional to the size of the area. In that case this 
formula can be written 
— JCP | oe 
7 (4m)3 % 2 
where ¢ is the azimuthal beam width, 7 the pulse 
length in seconds, ¢ the speed of light, and o is defined 
as the radar cross section of sea echo per unit area of 
the sea surface and hence is dimensionless. This quan- 
tity o is a function of many parameters: the state 
of the sea, the glancing angle of the incident beam 
(and therefore the range), the polarization, and the 
wavelength. A comprehensive program is under way 
in the Radiation Laboratory to check the assumptions 
underlying this formula and to determine the cross 
section o as a function of these parameters. 
Uhlenbeck has pointed out that the dependence of 
« on wavelength should be an especially sensitive 
function of the scattering mechanism assumed. For 
drops whose circumference is small compared to the 
wavelength, the scattering should be of the Rayleigh 
type, ie., varying as 1/A*. If one takes into account 
the lobe pattern of the incident field due to reflection 
on the water surface, the dependence is even faster, 
possibly as 1/A8. On the other hand, if we are dealing 
with reflection or diffraction from large curved sur- 
faces, then o should be substantially independent of 
wavelength or even increase with A. By measuring o 
simultaneously on two or more frequencies, it should 
be possible to decide between these mechanisms. 
Accordingly, such measurements were made in the 
summer of 1944 at Bar Harbor, using the calibrated 
S- and X-band mobile radars belonging to the Wave 
Propagation Group of the Radiation Laboratory. The 
site elevation was 1,500 ft, and the ranges about 
10,000 yd, so that the incident angles were quite small. 
The constants in the formula for P, were determined 
as accurately as possible. In addition, the power, pulse 
length, and beam width were made comparable in 
both systems. For relatively stormy sea conditions the 
ratio of c on the two wavelengths was found to be: 
og; 
