DIELECTRIC CONSTANT, ABSORPTION AND SCATTERING 297 
dental work on drop size was also done. The spread 
in drop size at a given time was rather large, probably 
because of the orographice character of the precipita- 
tion, which was made up of drops falling on the aver- 
age not much more than 1,000 ft, as compared to ordi- 
nary rain falling 5,000 to 6,000 ft. The longer period 
of fall in ordinary rain probably permits a greater 
number of drops to reach the characteristic sizes. 
ABSORPTION OF MICROWAVES BY 
THE ATMOSPHERE, BRITISH WORK' 
The working committee of the Ultra Short Wave 
Panel has presented a report in which the following 
questions are treated in detail: 
1. The absorption of microwave radiation by oxy- 
gen and vapor. 
2. The absorption of microwave radiation by water 
in macroscopic form, for example, rain, clouds, fog. 
3. The measurement in the laboratory of dielectric 
constants and conductivity which have already been 
described (see page 268 ). 
The work under item (1) consisted chiefly in show- 
ing that the attenuation given by Telecommunication 
Research Establishment radar experiments at Llan- 
dudno at a wavelength of 1.25 cm was consistent with 
the theoretical values given in reference 1. The experi- 
mental method consisted of comparing the echoes re- 
ceiyed on X- and K-band radars from a standard re- 
flector at short range and from land echoes at a large 
range. Care was taken to insure that the reflecting tar- 
gets did not reflect an amount of energy which was 
dependent on frequency. In this way a minimum value 
for the attenuation on X = 1.25 cm was found to be 
0.14 db per kilometer. New values based upon revised 
values of the widths of the various lines of water vapor 
and oxygen show that almost the whole of this atten- 
uation must be due to absorption by water vapor and 
further that to obtain an absorption as high as 0.14 
db per kilometer the frequency at the important water 
vapor line must lie close to 1.25 cm. 
The work under item (2) has been carried out, in 
the case of rain, by assuming what seemed a plausible 
distribution of drop sizes and calculating on this basis 
the attenuation that would occur at a standard pre- 
cipitation rate. Then by making a climatological sur- 
vey, information can be given of the proportion of 
time during which the attenuation can be expected 
to exceed a given value for a radar or radio commu- 
nication set working on a given wavelength at a given 
location. The calculations for a standard precipita- 
tion rate gave the following estimates for the maxi- 
mum attenuation likely to occur (that is, for the 
most unfavorable drop size distribution that was 
thought likely to occur). 
"By F. Hoyle, Ultra Short Wave Panel, Ministry of Supply, 
England. 
S band: 0.003 db per km per mm per hr. 
X band: 0.06 db per km per mm per hr. 
K band: 0.22 db per km per mm per hr. 
The climatological part of the program involves a 
great deal of statistical work and is not yet complete. 
The following preliminary results can be given. 
At Padang in Sumatra we may expect, twelve times 
a year, periods of 1 hr when the average attenua- 
tion on 1.25 em will be some 6 to 12 db per kilometer; 
on 3.0 em, 2 to 4 db per kilometer; and on 10 em, 0.1 
to 6.2 db per kilometer. 
In England only once a year will the average at- 
tenuation over a period of one hour reach 1 to 2.5 db 
per kilometer on 1.25 em; 0.3 to 0.8 per kilometer 
on 3.0 em; and 0.02 to 0.04 db per kilometer on 
10.0 cm. 
It should be noticed that these attenuation figures 
are for point-to-point communication and must be 
doubled in the radar case. 
DIELECTRIC CONSTANT AND LOSS 
FACTOR OF LIQUID WATER AND 
THE ATMOSPHERE® 
In propagation problems the knowledge of the 
electric properties of the ground, sea, and fresh water, 
as well as those of the atmospheric gases, are of funda- 
mental importance. Collected here are the available 
data on these materials. Clearly, the study of the 
reflection coefficients leads indirectly to the dielectric 
properties of these materials. Here we will be con- 
cerned more with the direct determination of their 
dielectric constants and loss factors. 
Experimental Methods 
RerLucrion-TRANSMIssIon MrtHop 
First we should like to sketch here the basis of the 
different experimental methods used in the determina- 
tion of the dielectric properties of materials. 
One of these is the reflection-transmission method 
which was used by Ford*’ in his studies of the prop- 
erties of the ground. This same method has also been 
used recently by Saxton®®*® on water in investigating 
the temperature dependence of its dielectric properties 
at 1.25 and 1.58 em. 
If the power transmitted through two thicknesses 
d, and d, of the material in question are 
P, = Po e-eh ’ 
P2 = Po e7%4, 9 
then the absorption coefficient a is given by 
2.3 P, 
= = 77 
R= Gh logio P,’ (77) 
a 
*By L. Goldstein, Columbia University Wave Propagation 
Group. 
