RADAR STORM DETECTION 
is detecting those portions of the storm which give 
relatively weak echoes. 
Attenuation 
Attenuation of microwaves is caused by absorption 
of oxygen and water vapor in the atmosphere and by 
absorption and scattering from cloud and precipitation. 
The water vapor molecule has an electric dipole moment 
which interacts with the electric field of the radiation 
and has resonance at 1.33 em. A stronger absorption 
band at a wave length of less than 0.2 cm also con- 
tributes to the attenuation. The oxygen molecule has 
a magnetic dipole moment which interacts with the 
magnetic field of the radiation with resonance at 0.5 
em. Because of the uncertainty of the line breadths of 
the absorption bands, actual values of the oxygen 
attenuation are known only for certain regions. Table 
II gives Van Vleck’s [12, 13] values of the attenuation 
Tasie I]. ArrenuatTion* Dur To OxyGEN 
peer enenetn Oxygen Water vapor 
1 0.014 —0.036 0.071 
1.25 0.010 -0.026 0.21 
3 0.0072-0.018 0.0046 
10 0.0066-0.014 0.0031 
*TIm decibels per kilometer at 760 mm pressure and 10 g 
kg— water vapor, and at a temperature of 293K. 
due to oxygen and water vapor. The attenuation due 
to oxygen is directly proportional to the atmospheric 
pressure. Van Vleck favors the lower values of oxygen 
attenuation. At a wave length of 3 to 10 cm the total 
atmospheric attenuation is therefore 0.01 to 0.02 db 
km. 
More serious is the attenuation due to rain. For mod- 
erate to heavy rains, computed on the basis of the mean 
drop-size distribution of Laws and Parsons, Ryde’s 
values for wave lengths of 1.25, 3, and 10 cm are 0.14, 
0.023, and 0.003 db km“ mm hr!, respectively. 
Values of theoretical attenuation given by Haddock in 
an unpublished paper differ somewhat from Ryde’s 
values. 
Considerable error is involved in experimental meas- 
urements of rain attenuation since intensity and drop- 
size distribution vary over the path. In experimental 
determination of the attenuation at 0.62 em by Mueller 
[9] there was excellent agreement with Ryde’s theoreti- 
cal values. Measurements by Robertson and King [10] 
indicate a scatter of values and give a mean of 0.03 
db km mm~ hr at 3.2 em. This is much higher than 
the average theoretical values but within the theoreti- 
cal limits. For 1.25 em, Anderson and collaborators 
{1] found an average attenuation of 0.18 db km 
mm! hr~! for Hawaiian rainfall. This exceeds Ryde’s 
theoretical maximum values. Reasons for this discrep- 
ancy are difficult to see especially since Anderson’s 
experiments appear to be the most careful. 
Vertical Structure of Precipitation 
Precipitation as observed by radar may be generalized 
into two categories: showers and continuous precipita- 
1285 
tion. Continuous precipitation is associated with small 
vertical velocities of the order of 10 to 30 em sec7!, 
while the vertical velocities of showery precipitation 
may exceed 1 m sec™!. There is, of course, a continuous 
transition between the two. On an RHI scope,’ con- 
tinuous precipitation is characterized by a horizontal 
layer whereas showers appear as elongated vertical 
bands. The two types may occur within a few miles of 
each other and may be found mixed in almost any 
proportion. 
A detailed study of the vertical structure of precipi- 
tation has been made by Langille, Gunn, and Palmer 
[5]. By calibrating the gain of the radar and by assum- 
ing that the summation of the sixth powers of the 
diameters of the drops is proportional to the square of 
the mass of water illuminated by the beam, they were 
able to determine lines of equal water content. Some 
conclusions from these and other observations of the 
two extremes of precipitation are: 
1. Showers 
a. Within the space of a mile the liquid water content 
may change by a factor of more than 100. 
b. The liquid water content may be greater aloft 
than near the ground. 
c. Heavier precipitation is generally associated with 
greater vertical extent of the storm. 
d. The motion of the dense portion of a storm (high 
liquid-water content) is always towards the ground. 
2. Continuous Precipitation 
a. A bright band, a horizontal layer of high echo 
intensity 100 m to about 300 m below the freezing level, 
is an identifying feature of continuous rain. (A sepa- 
rate section below is devoted to the theory and descrip- 
tion of this phenomenon in relation to the vertical 
structure of continuous precipitation.) 
b. Radar echoes along the horizontal change less 
rapidly than echoes from showers. 
c. The maximum height of precipitation echoes often 
remains constant while the rain intensity varies. Such 
behavior has been observed on both K- and X-band 
radar with the antenna directed towards the vertical. 
The maximum height of the radar echo indicates the 
level at which particles first approach precipitation size 
(greater than 0.1 mm) and may vary with the radar. If 
the radar echo dies off gradually until it is indistinguish- 
able from the noise level, there is no indication as to 
where the actual top of the cloud may be; but if the 
echo falls off suddenly (except near the bright band), 
then it is reasonable to expect that the visual top of 
the cloud is not far above. The latter type of echo is 
associated with precipitation from well-developed 
cumulonimbus clouds, while the former is associated 
with stratiform clouds and represents a very gradual 
growth of the precipitation elements. 
The maximum height of radar echoes observed in 
temperate latitudes is found to be 15,000 to 20,000 ft 
for most widespread winter rains and near 30,000 ft 
for precipitation from cumulonimbus clouds in summer. 
The maximum height observed in a thunderstorm at 
3. RHI = Range Height Indicator. The abscissa is hori- 
zontal range, the ordinate is usually angle of elevation. 
