TEMPERATURES AND PRESSURES IN THE UPPER ATMOSPHERE 
By HOMER E. NEWELL, Jr. 
Naval Research Laboratory 
Introduction 
The stratosphere was discovered in the pioneering 
days of high-altitude balloon research. Between 1899 
and 1902 experiments performed by Assmann and Teis- 
serenc de Bort revealed a cold isothermal layer extend- 
ing upwards from the top of the troposphere. The 
initial height of the layer appeared to vary from about 
6 km in the polar regions to 18 km above the equator. 
As a result of such balloon measurements, it was long 
supposed that beyond the troposphere the temperature 
ceased to change with altitude. In the presence of the 
earth’s gravitational field such an isothermal region 
would, of course, be characterized by a diffusive sepa- 
ration of its various constituents, with the heavier gases 
settling out and the lighter ones predominating at the 
higher altitudes. The name stratosphere aptly describes 
such an atmosphere. 
In the course of time, however, evidence began to 
accumulate to show that the temperature of the atmos- 
phere is not the same at all heights above the tropo- 
pause. On the basis of data now available from sound 
and meteor studies, from rocket measurements, and 
from a number of other sources, it is plain that atmos- 
pheric temperatures vary markedly with altitude. In 
higher latitudes and the polar regions an isothermal 
region does exist above the troposphere; but as one 
moves southward the purely isothermal stratosphere 
disappears to be supplanted by a rather flat temperature 
minimum somewhere between 10 and 20 km above the 
surface of the earth [16]. Above the (improperly named) 
stratosphere, in the region now referred to simply as the 
upper atmosphere, the air is definitely not isothermal. 
On the other hand, the total pressure difference between 
ground and 110 km is nearly that to be expected in an 
isothermal atmosphere of about 240K.! In this sense 
one may regard 240K as an average temperature for the 
entire region below 110 km. Above 110 km there are 
some indications of very high temperatures. 
There are now numerous sources of data on which to 
base conclusions about temperatures and pressures in 
the upper atmosphere. Some of these are discussed in 
the sections below. No attempt is made, however, 
either to exhaust the literature or to furnish complete 
details of the temperature and pressure studies which 
are discussed. 
Balloon Studies in the Upper Atmosphere 
Until recently the highest-flying balloons could rise 
not much higher than 30 km. Measurements on such 
1. This statement is based on the pressure curve of Fig. 7 
below. 
balloon flights indicated a small but definite rise in the 
temperature near the top of the stratosphere. Such a 
rise appeared consistent with temperatures deduced 
from studies of the anomalous propagation of sound, 
which suggested a sharp rise above 30 km, but which 
furnished only an indirect determination of temper- 
ature. Now, with improved balloons, it is possible to 
trace out by direct measurement an additional 10 km 
or more of the temperature curve. 
The graph of Fig. 1 shows a curve of temperature 
variation from the ground up to 43 km. The measure- 
MINUTES AFTER RELEASE 
50 100 150 
(KM) 
ALTITUDE 
220 240 260 
TEMPERATURE(°K) 
280 
Fie. 1—Results obtained from balloon flight at Evans 
Signal Laboratory, Belmar, New Jersey, on September 28, 1948, 
1:20 p.m. The length of the wind vector is proportional to the 
wind speed (mph). The tail of the wind vector indicates the 
direction from which the wind was blowing. Thus, near the 
ground, the wind was northeasterly; at 15,000 feet, northerly; 
at 40,000 feet, westerly; and at 50,000 feet northwesterly (taken 
from [4] Brasefield: Sci. Mon., 68:398 (1949), by permission 
of the publishers). 
ments were made at Belmar, New Jersey, on September 
28, 1948 at 1:20 p.m., using a new type of balloon 
especially developed for the Signal Corps [4]. It will be 
noted that a positive temperature gradient is observed 
from 15 km to the peak of the flight. The measured 
temperatures indicate an abrupt inversion in the neigh- 
borhood of 15 km, and show no isothermal region. They 
do not even appear consistent with any flat temperature 
minimum in the stratosphere. The measurements are 
303 
