INSTRUMENTS AND TECHNIQUES FOR METEOROLOGICAL MEASUREMENTS 
To improve the accuracy of wind measurements 
above 80,000 ft, it will be necessary either to improve 
the accuracy of pressure measurement at these altitudes 
and take advantage of the rather uniform ascensional 
rates of the balloon, or to resort to transponder techni- 
ques and obtain accurate slant ranges. It is still too 
early to determine which method is more economical. 
Meteorological Recorder. As described above, the 
properly shaped pulses containing the temperature, 
pressure, and relative humidity data are fed to a fre- 
queney meter which converts the incoming frequency, 
over the range from 8 to 200 eps, to d-e voltages lin- 
early proportional to frequency. These voltages are 
measured by a modified Leeds and Northrup-type re- 
cording potentiometer. The unit includes circuits for 
automatic range adjustment. Drifts occurring in the 
airborne transmitter are automatically corrected by 
this adjustment. The equivalent frequency is printed 
on a 10-in. strip chart that covers the range from 0 to 
200 eps. The chart can be read within 0.1 of a division, 
producing an error in temperature of about -t0.2C. 
Miscellaneous Accessories. In order to improve the 
reliability of base-line checks, a portable chamber has 
been designed to maintain reasonably constant values 
of temperatrue, pressure, and relative humidity. These 
quantities are known within 0.1C, 14 mb pressure, and 
1 per cent relative humidity. A circular evaluator that 
permits the simultaneous lock-in of both temperature 
and relative humidity is also available. This important 
innovation was made possible by recognizing that the 
shapes of the resistance-humidity curves of the humid- 
ity strips are nearly identical. The dew-point tempera- 
ture for any relative humidity may be read directly off 
the scales. By use of this evaluator the efficiency and 
accuracy of humidity computations have been signifi- 
cantly mereased. 
General Remarks. In appraising the over-all per- 
formance of this radiosonde-radiowind system, it is 
important to bear in mind that the design features are 
such that no individual calibration of the radiosonde 
after leaving the factory is necessary, and furthermore 
that each of the components can be mass produced— 
on the order of 100,000 per year. The manufacturing 
tolerances placed on the individual components are 
indeed quite severe. If the radiosondes were to be used 
for research purposes, where individual calibration of a 
small number could be justified, the over-all accuracy 
of the measurements would be greatly improved—the 
temperature error reduced to approximately +0.2C, 
the pressure errors to --0.5 mb, and the wind errors 
for altitudes above 40,000 ft to about one half (or less) 
of those computed in Fig. 7. 
Because of the uncertainties of the solar radiation 
correction, the effects of moisture condensation, and 
the lag, it does not appear that temperature measure- 
ments with the present thermistors can be made with 
probable errors less than --0.2C up to altitudes of 
100,000 ft. Very little data are available on the ac- 
curacy of temperature measurements within clouds, 
using the exposed thermistors. The problem has not 
received the attention that its importance merits. 
1213 
In order to improve the pressure measurement at 
high altitudes, serious consideration should be given 
to the use of the hypsometer. Preliminary measure- 
ments on hypsometers, designed for radiosonde use, 
indicate that in the range of from 50 mb to 1 mb an 
accuracy of about 0.1 mb should be expected. Further- 
more, the use of the hypsometer for the entire pressure 
range is feasible, provided pressure measurements at 
1000 mb to within 7 mb are acceptable. Pressure meas- 
urements at the lower pressures would be made with 
the same percentage error. Some modification of the 
meteorological recorder would also be necessary. 
The problem of humidity measurements in the free 
atmosphere is indeed the most difficult one. As out- 
lined above, the three major objections to the present 
lithium chloride strip are (1) the deterioration of the 
element on direct exposure to water droplets, (2) the 
very high lag constant at low temperatures, and (3) 
the insensitivity at very high humidities. Although 
various salts have been suggested as substitutes for 
lithium chloride to improve the performance of these 
elements, it does not appear that the use of electrolytes 
will overcome the basic difficulties of low-temperature 
hygrometry and of irreversible dilution when exposed 
to water droplets. One encouraging development on 
the horizon is the possible use of a carbon humidity 
strip. Although very little information is available at 
present, it does appear that the carbon elements have 
lag constants of the order of 0.4 sec at 25C; further- 
more they are not subject to the dilution difficulties 
of the electrolytic strips, and are quite sensitive to 
humidity changes at high relative humidity. The op- 
portunities for research in hygrometry are still very 
great. 
Radar-Wind and Radarsonde Systems. In the pre- 
ceding paragraphs a description has been given of a 
wind measuring system that required a knowledge of 
azimuth angle, elevation angle, and height. Another 
system in use, capable of accurate wind measurement, 
employs a microwave radar. Measurement is made by 
automatically tracking a balloon-borne passive target, 
in the form of a corner reflector, with a suitable ground 
radar set operating at either 3000 or 10,000 me sec. 
Measurements are then made of azimuth angle a, 
elevation angle 6, and slant range r. In addition, the 
time derivatives of these quantities, a, 6, 7, are avail- 
able. With this information, wind velocity can readily 
be computed. The accuracy of the angular measure- 
ment is comparable to that of the direction finders; 
however, the accuracy in range measurement is greater 
than the corresponding accuracy in height determina- 
tion, especially at high altitudes. Furthermore, in the 
radar system the cosine of the elevation angle is used 
to compute horizontal distance; in the radio direction- 
finding system, the cotangent is used. For low-elevation 
angles, therefore, the radar system is capable of greater 
accuracy in wind measurement than the radio direc- 
tion-finding system. The possible limitation of this 
system for wind measurement is that of range. However, 
by replacing the passive target with an active trans- 
ponder or beacon, this limitation can readily be over- 
