1286 
Belmar, New Jersey, was 50,000 ft and in Florida it was 
observed by Byers to be 55,000 to 60,000 ft. Precipita- 
tion that reaches the ground as snow generally gives a 
lower maximum echo height than rain, and snow show- 
ers have been observed with maximum heights lower 
than 5000 ft. 
The temperature at the top echo of most mature 
thunderstorms is generally below —30C. It is at about 
this temperature, according to W. and HW. Findeisen 
[3], that the number of freezing nuclei become numer- 
ous, with large vertical velocities and rapid growth in 
the high liquid-water content of the thundercloud. How- 
ever, small thunderstorms have been observed with 
the temperature at the top echo and the visual top at 
about —18C [7]. 
It has been observed that rainfall with maximum echo 
heights corresponding to temperatures above —9C 
rarely reaches the ground. The smaller number of ice 
particles produced at such high temperatures would 
grow rapidly to precipitation size, drop out of the cloud, 
and evaporate in the unsaturated air beneath. With 
greater vertical extent of the cloud above OC, more 
numerous ice nuclei are produced which grow to pre- 
cipitation size over a larger height interval and are 
numerous enough below the cloud to maintain near- 
saturated conditions. 
Observations of twelve thunderstorms in New Mexico 
indicate that the first echo from a growing thunderstorm 
appears at about the —10C isotherm with the top of the 
cloud at about —30C. The echo then rises rapidly to a 
height where the average temperature is —30C with 
the visual top of the cloud at about —32C. Although 
visual observation of the top is difficult in a degenerat- 
ing thunderstorm it appears that both the top of the 
echo and of the cloud descend thereafter. 
The Bright Band and Precipitation Growth 
A characteristic A-scope presentation of the bright 
band is shown in Fig. 3. The first intense signal is the 
transmitted pulse; above that, the weaker echoes from 
Fic. 3—Bright band observed at Belmar, N.J., January 26, 
1949 by Weinstein [14]. The markers are at 1-mile intervals. 
The ordinate represents signa] intensity. 
RADIOMETEOROLOGY 
the precipitation are fairly constant in intensity up to 
the bright band, the intense saturated signal at about 
7500 ft. The width of the bright band appears to be less 
than a thousand feet. Above the bright band the signal 
falls off gradually and disappears at about 15,000 ft. 
Observation of the bright band was first reported by 
the Canadian Army Operational Research Group in 
1945. From twelve days of observation it was concluded 
that the echoes from the bright band were from just 
below the OC isotherm. It was observed only in strati- 
fied clouds although some vestiges of the bright band, 
considerably weaker in contrast, have been detected in 
the final stages of a thunderstorm. In some cases the 
bright band was observed in some sectors but not in 
others. 
Subsequently, aircraft investigations revealed that 
immediately above the bright band the precipitation 
was in the form of snow, in the bright band it was 
melting snow, and below the bright band it was rain. 
As an explanation it was surmised that “a melting snow- 
flake is covered by a film of water which causes it to act 
as a raindrop effectively larger than the size of the 
raindrop resulting from the complete melting of the 
snowflake.” 
Ryde [11] rejected this explanation and offered an 
alternative theory. At temperatures between —4C and 
OC a cloud of individual ice crystals aggregate to form 
snowflakes, thereby giving increased reflectivity. As a 
result of this aggregation, the relative intensity was 
estimated to increase from 1 to 40. As the snowflake 
melts, the reflectivity is further increased owing to the 
higher dielectric constant of water as compared to ice. 
Because of this effect, a raindrop resulting from the 
melting of a snowflake has about five times the reflec- 
tivity of the original flake. The decrease in signal 
strength below the bright band was explained as due 
to the increased fall velocity of raindrops as compared 
to snow; hence the number of raindrops per unit volume 
of air is smaller than that of snowflakes above. Since 
the fall velocity of raindrops is about five times that of 
snowflakes, the signal strength would be one-fifth that 
of the melting snow. 
From the theory of storm detection it may be seen 
that the reflectivity of a melting snowflake is propor- 
tional to K?/pV where K = (e — 1)/(€ + 2); p is the 
density of the mixture, and V is the fall velocity. 
Ryde assumed that 
eee pe (6) 
p 
Pi Pw 
where subscripts 7 and wrefer to ice and water, respect- 
ively, and m; and m,, are the respective percentages of 
the total mass of the snowflake (m; + m» = 1). This 
equation is true for a heterogeneous mixture of ice and 
water but it is not true where the water is on the out- 
side of the snowflake. A particle with an inner core of ice 
and an outer shell of water has a dielectric constant more 
nearly that of the water than equation (6) would indi- 
cate because of the attenuation of the radio wave towards 
the interior of the particle (as in the “skin effect”). The 
increase in signal strength at the top portion of the 
