AIRCRAFT METEOROLOGICAL INSTRUMENTS 
By ALAN C. BEMIS 
Massachusetts Institute of Technology 
INTRODUCTION 
The first scheduled weather flights were made in the 
early 1930’s and were essentially vertical soundings. 
The airplanes carried simple meteorographs which re- 
corded temperature, pressure, and humidity in much 
the same manner as the balloon-borne meteorographs 
of that period. Radiosondes being as yet undeveloped, 
the use of airplanes was the most practical method of 
obtaining vertical measurements promptly for fore- 
casting purposes. At present the airplane, with its 
greater speed and cruising range, is more commonly 
used for measurements aloft along a chosen horizontal 
line, and it is for flights of this type that most instru- 
mentation is now designed. 
Much of the instrumental development to date has 
been carried out in connection with highly specialized 
research programs of various sorts. The study of air- 
craft icing, for example, has required an extensive 
program of instrumentation, much of it designed for 
important meteorological quantities never before meas- 
ured in the free air [4, 27, 41, 52]. For such a program 
it 1s often practicable to use very complex laboratory 
instruments, expensive to build and operate. The author 
believes we now suffer from a serious lack of simple, 
well-engineered instruments, which might be relatively 
inaccurate, but nevertheless of great value when com- 
monly available. For example, the Thunderstorm Proj- 
ect [5] operating in 1946-47 was unable to make any in- 
flight measurements of liquid-water content because 
the only available instruments for the purpose were 
still experimental. Even temperature measurement was 
a problem, particularly in the presence of liquid water. 
In short, aircraft meteorological instrumentation was 
then and still is im the laboratory. 
A more vigorous attack on these problems is recom- 
mended. Weather reconnaissance by aircraft is an ex- 
pensive and important operation, yet relatively small 
sums have so far been spent on instrumentation for 
those aircraft. Though many important meteorological 
quantities are best observed visually, others require 
instruments, preferably recording instruments. There 
is need here for inventive genius, good engineering, 
and funds for development work. 
THE PROBLEM 
For any observation to be useful it must be located 
in time and space. Let us assume that location in time 
is no problem and that ordinary navigational aids will 
locate the measurements in a horizontal plane with 
sufficient accuracy. We may also assume that some form 
of radio altimeter will measure true altitude with suf- 
ficient accuracy to give significance to local barometric 
pressure readings. Then the measurable quantities 
which are most important to the meteorologist are 
temperature, humidity, pressure, wind velocity, and 
liquid (or solid) water content. Also of considerable 
importance are droplet or snowflake size, turbulence 
and drafts, and electric field strength. Most of these 
quantities can be either measured directly or calculated 
from several related quantities, but are specified above 
in something approaching fundamental parameters. 
More precisely, the basic quantities referred to are: 
1. Air temperature T, in degrees centigrade. 
2. Humidity q, in grams of water vapor per kilogram 
of moist air. 
3. True altitude h, in meters above mean sea level. 
4. Air pressure p, in millibars. 
5. Wind velocity V, referring to both the direction in 
degrees and the speed in knots. 
6. Liquid (or solid) water content MM, m grams per 
cubic meter. 
All of the above quantities can be measured with 
relative ease at the ground. Most of the difficulties 
aloft stem directly or indirectly from the high velocity 
of the measuring ‘‘platform.” Problems of instrumental 
lag, accurate location of the pomt of measurement, 
and automatic recording are far more acute, and new 
problems involving kinetic energy arise. Aircraft veloc- 
ities are about ten times greater than the wind veloc- 
ities and raindrop fall velocities encountered at a ground 
station. Quantities involving energy transfer are then 
one hundred times greater. This means that adiabatic 
heating effects, and difficulties caused by the presence 
of liquid water, are so severe as to be unsolved for some 
conventional instruments. These problems become par- 
ticularly acute in measuring temperature, and will be 
treated at some length in the next section, where sev- 
eral new and promising methods are also described. 
MEASUREMENT OF TEMPERATURE 
The measurement of temperature from aircraft at 
speeds over 100 knots is seriously complicated by dy- 
namic heating of the thermometer element. The heating 
is partly due to adiabatic compression of the air and 
partly to friction. It differs, therefore, for each type 
of thermometer housing, or lack of housing, and for 
various mounting locations on the aircraft, but is gen- 
erally proportional to the square of the true air speed. 
This source of error has received careful study [21, 
29]. Theoretical and empirical correction factors as 
functions of air speed are available for the most com- 
mon thermometer types and are reasonably satisfac- 
tory in dry air [47]. Some engineers have even designed 
built-in compensators working from air-speed devices. 
A more reasonable approach has been made by others 
[44, 51] using expansional cooling in the housing to 
1223 
