HIGH LAYERS OF THE EARTH’S ATMOSPHERE\*

E. O. Hulburt

Submitted 1948 | SovietRxiv: ru-194801.62148 | Translated from Russian

Full Text

HIGH LAYERS OF THE EARTH’S ATMOSPHERE*

E. O. Hulburt

In recent years interest in the high layers of the earth’s atmosphere has grown rapidly, although during the war period many investigations were classified or entirely abandoned. Now a powerful stimulus has appeared in the form of new apparatus, such as the V-2 rocket, capable of reaching considerable heights and thereby making possible direct experimentation at great altitudes. Here I propose briefly to describe current information about the atmosphere, up to the ionosphere, and to note the substantial contribution that has already been obtained with the aid of V-2 rockets.

DIRECT OBSERVATIONS

Direct information on the temperature, pressure, composition, and motion of the air at great heights has been obtained with the aid of aerostats, both with observers and without them, and also from observations of meteor trails and the smoke from bursts of artillery shells. The solid curves in Fig. 1 show the change of temperature with altitude up to a height of 20 km; they were obtained with radiosondes in Europe, at about 48° north latitude. The summer curve is the average of the records of 231 flights, and the winter curve is the average of the records of 185 flights. It is not specified whether these refer to daytime or nighttime conditions. Some data relating to greater heights and shown in Fig. 1 by a dashed line were obtained by the U.S. Weather Bureau from radiosonde flights that took place near sunset in Omaha (Nebraska, 41° north latitude). Complete daytime data for the higher layers of the atmosphere are still lacking, owing to the difficulty of determining the temperature of the surrounding air in full sunlight. Likewise, complete data for nighttime are also lacking.

From Fig. 1 it is evident that, for middle latitudes, the air temperature decreases with altitude by approximately 6° C per 1 km, up to heights of 10–15 km, where a temperature from −40 to −70° C is reached.

* E. O. Hulburt, Journ. Opt. Soc. Amer., 37, 405 (1947). Translated by G. V. Rozenberg.

At great heights in winter, near sunset, the temperature remains almost constant or increases slightly up to heights of about 24 km. In summer, near sunset, there is a considerable increase in temperature in the altitude interval approximately from 20 to 34 km. Direct measurements of air temperature above 35 km are entirely lacking.

Fig. 1. Temperature according to balloon-sonde data.

Fig. 1. Temperature according to balloon-sonde data.

The region of the atmosphere approximately from 10 to 20 km, where the temperature changes little with height, is called the “stratosphere” or the “isothermal region.” The region situated above it is called the “high layers of the atmosphere.” The stratosphere varies with latitude; its lower boundary is located at about 10 km altitude at 60° north or south latitude in summer and at about 15 km at the equator. The corresponding temperatures are approximately −45 and −70° C. For middle latitudes, when the barometer readings are low, the stratosphere is 5–10° warmer than when they are high. Nothing is known about the diurnal variations of the temperature and height of the stratosphere.

Determinations of wind speed on the basis of balloon-sonde observations show that the speed increases with height until it reaches-

reaches the surface of the stratosphere. Here the velocities decrease and then, with increasing altitude, begin to rise again. Average monthly data² for times near sunset, obtained with the aid of sounding balloons flown in the region of Omaha, are shown in Fig. 2. From observations of smoke from the bursting of artillery shells at an altitude of 30 km, carried out over southeastern England, Johnson³ found that the average summer and winter wind velocities were respectively 43 and 133 km per hour; these values are marked in Fig. 2 by crosses. The maximum value—236 km per hour—was recorded in winter. The drift velocity of meteor trails was measured in 26 cases⁴. Winds and vortices with velocities above 80 km per hour were noted, and in two cases about 240 km per hour. Meteor trails were observed at levels of approximately 40 to 110 km. Immediately after their formation all the trails became curved and twisted, as if by atmospheric vortices and turbulence.

Fig. 2. Wind velocity. Curves—data from sounding balloons; crosses—data from observations of smoke from bursts of artillery shells.

Fig. 2. Wind velocity. Curves—data from sounding balloons; crosses—data from observations of smoke from bursts of artillery shells.

As regards the composition of the atmosphere, from air samples obtained at great heights during balloon flights, the amount of oxygen was determined⁵ up to an altitude of 29 km, and helium up to an altitude of 21 km. It was found that at all the altitudes reached the proportion of the corresponding gases by volume remains constant, with an accuracy of a few percent, and corresponds to their proportion at sea level.

Regener’s data for oxygen and Paneth and Glückauf’s data for helium are shown in Fig. 3. However, quite real small and variable deviations from constancy were observed; their occurrence may be attributed to irregularities in the lower atmosphere, such as weather.

Fabry and Buisson⁶ discovered that ozone is present in the upper layers of the atmosphere and that its total amount in a vertical column of air corresponds to a layer 2 to 4 mm thick under normal conditions.

International measurements in 1926—1929, carried out under Dobson’s direction7, found that the amount of ozone increases with latitude from 2 mm at the equator to approximately 3.6 mm at 68° northern latitude and 44° southern latitude. The ozone content varies with the seasons, having a maximum in spring and a minimum in autumn. It cannot yet be said with certainty whether diurnal variations in ozone content occur over a given point. Some data7 indicate that diurnal variations, if they do occur, are insignificant.

Figure 3. Oxygen and helium according to balloon-sonde data.

Fig. 3. Oxygen and helium according to balloon-sonde data.

Figure 4. Vertical distribution of ozone. Solid curves—balloon-sonde data; dashed curves—the “Umkehr effect.”

Fig. 4. Vertical distribution of ozone. Solid curves—balloon-sonde data; dashed curves—the “Umkehr effect.”

Curves of the vertical distribution of ozone are shown in Fig. 4. The solid curves correspond to direct observations with the aid of balloon sondes8; the dashed curves correspond to indirect measurements based on the “Umkehr effect”9. The numbers beside each of the curves indicate the total amount of ozone in the vertical column corresponding to the given case. It is evident from the figure that above

At about 25 km the ozone content decreases with height, and the distribution of ozone above 30 km is not sufficiently well known. The ozone content at higher levels is of great importance for the temperature at these heights.

THEORETICAL CONSIDERATIONS

The theoretical analysis of the atmosphere had a twofold significance: (a) the “understanding” or “explanation” of characteristics such as composition, pressure, temperature, etc., observed in the lower layers of the atmosphere, as a consequence of known or presumed causes; and (b) the extrapolation of the theory of the lower layers of the atmosphere to its upper layers, for which direct observations are lacking, with the aim of determining or predicting the unknown characteristics of these layers.

Probably the most complete works of this kind were published in 1928 by Maris and Gouzen1. Some of their conclusions have now been confirmed, while others require modification in the light of later investigations, which supplemented or changed various data used in these works. The physical essence of the theory is given below.

The atmosphere is a mixture of gases held at the surface of the Earth by the force of gravity. If there were no wind and the temperature were constant, the atmosphere would be in a state of isothermal equilibrium; each gas would be distributed with height according to the gas laws and the force of gravity, and the proportion of light gases would increase with height. If the atmosphere were uniformly mixed, for example by wind, the proportion of all gases would be unchanged at all heights.

Let us now suppose that the atmosphere has a temperature constant in time, that it is uniformly mixed in the vertical direction, and that it is free of wind, while diffusion and gravity act, tending to restore the conditions of isothermal equilibrium. The lighter gases diffuse upward, the heavier ones downward. The rate of diffusion increases with height, as a result of which the isothermal distribution is attained more rapidly in the higher layers and more slowly in the lower layers. The level at which the region of complete diffusion comes into contact with the region of homogeneous mixture is called the “diffusion level.” Maris showed that, for example, for helium at 0° C, the diffusion level is located at 150 km after 0.53 day; after one day it descends to 146 km; after two days, to 140 km; after five days, to 136 km; after a year, to 110 km. It was shown that the diffusion level depends on the nature of the gas and on the temperature, being, for helium, respectively at −50 and +100° C, approximately 12 km lower and 25 km higher than at 0° C. Now, although the temperature and the mixing capacity of the wind at great heights are not known exactly, it is nevertheless known that winds occur at least up to 100 km above sea level, and it seems probable that the tem-

temperature does not go appreciably beyond the limits of $-50$ to $+100^\circ$C. Therefore it seems a sufficiently reliable conclusion that the diffusion level is situated no lower than 100 km for all gases entering into the composition of the atmosphere. This means that, apart from the exceptions discussed below, the structure of the atmosphere remains approximately unchanged from sea level to heights of the order of 100 km. This is the viewpoint now generally accepted. The data of Fig. 3 clearly confirm this conclusion: the dashed curves giving the proportions of O$_2$ and He corresponding to a diffusion level located at about 10 km illustrate the fact that in reality the diffusion level lies above 30 km.

The exceptions with respect to the constancy of the composition of air with height are water vapor, ozone, atomic oxygen, atomic nitrogen, and nitrogen oxides. The first two cannot significantly alter the composition of the air, since the total quantity of water vapor in a vertical column of air with a cross section of 1 cm$^2$ cannot exceed approximately 2.6 g. This quantity is calculated on the assumption of saturated water vapor throughout the whole path upward. The total quantity of ozone cannot exceed approximately 4 mm. This quantity is small in comparison with 8.3 km of a vertical column of air reduced to normal conditions.

A detailed theoretical discussion of atomic oxygen, atomic nitrogen, and nitrogen oxides lies beyond the scope of the present article. It will suffice to indicate that, owing to the probability of the entry into the atmosphere of the Sun’s short-wave ultraviolet radiation, it seems beyond doubt that molecules of oxygen and nitrogen dissociate, forming nitrogen oxides. But to what extent and at what height these phenomena take place still remains in the realm of conjecture.

Theoretical calculations of the temperature of the high layers of the atmosphere were based on the absorption of solar and terrestrial radiation by atmospheric gases. The spectrum of the Sun was extrapolated into the region shorter than 2900 Å—the region that remained unknown—by assuming that the spectrum of the Sun coincides with the spectrum of an absolutely black body at 6000°K. An important role in regulating the temperature of the atmosphere was ascribed to carbon dioxide, water vapor, ozone, and oxygen, because their absorption bands are situated in the corresponding region of the spectrum. Of these, the most important is ozone, since it is precisely to it that the ultraviolet absorption band extending from 2900 to 2300 Å owes its origin.

All the absorption bands of each of the gases participate in the absorption of the energy of the Sun and the Earth during the day, and the infrared bands in the radiation of energy at night. The resulting temperature depends on the ratio between the intensities of radiation and absorption of energy; and these, in turn, depend to a large extent on the vertical distribution of ozone.

Assuming a definite distribution of atmospheric gases with height and estimating the mixing wind, Möris$^{10}$ calculated that

on a summer day at latitude 50° the temperature increases with height approximately from −40° C at 30 km to +90° C at 100 km, while on a winter night the temperature remains practically the same—about −40° C—at all levels above 30 km. Gowen^10 obtained results broadly similar to those of Méric. He also calculated the influence of assumed variations in the amount of water vapor and ozone. His calculations were based on the assumption that the maximum of the vertical distribution of ozone lies near 50 km, which in 1928 was considered to correspond to reality. Later observations showed that this assumption is erroneous and that the ozone maximum is located between 20 and 30 km. The more recent investigations mentioned below have established that solar radiation in the ultraviolet absorption band of ozone between 2300 and 2800 Å is substantially less than that of a black body at a temperature of 6000° K. Therefore the quantitative results obtained on the basis of theoretical considerations are subject to revision. Such a revision, as well as the further development of the theory, may have to be postponed until new information on the solar spectrum and the upper layers of the atmosphere makes such an excursion into theory worthwhile.

In conclusion it may be said that the theoretical consideration of the upper layers of the atmosphere has led to two results: (a) the composition of the atmosphere is almost the same in the altitude interval from 0 to approximately 100 km, and (b) somewhere above 30 km the atmosphere becomes warmer, but precisely where this occurs, and how strongly the temperature changes from day to night, could not be obtained from the data known up to 1946.

It should be noted that cosmic rays remain outside the discussion because none of the essential characteristics of the upper layers of the atmosphere can as yet be connected with their effect. It seems quite possible, however, that in these layers chemical or photochemical reactions will be discovered which depend on high-energy radiation or owe their origin to it.

INDIRECT DATA

If the structure of the atmosphere is known at all altitudes, then for all altitudes the relation between pressure and temperature is also known and, consequently, the pressure can be calculated from the temperature, and conversely. Four kinds of phenomena may be named that lead to conclusions about the temperature and pressure in the upper layers of the atmosphere: 1) the propagation of sound, 2) atmospheric tides, 3) the brightness of the twilight sky, and 4) meteors.

In the case of sound waves from a sufficiently intense source, for example from an explosion or an artillery salvo, it was observed^11 that at distances of 100 to 200 km the sound arrived substantially later than the time that would have been expected from the distance and the known speed of sound. Only a small fraction of this anomalous

may be attributed to the wind over an anomalously long propagation time. To explain the long duration of sound propagation, it is assumed that the sound ray penetrates high into the atmosphere and bends back toward the earth. This assumption is supported by the fact that often the zone of silence, extending from 50 to 100 km from the source, is surrounded by a zone of audibility, and that in the zone of audibility the sound ray arrives at an angle to the horizon.

Fig. 5. Temperature according to sound-wave propagation data.

Fig. 5. Temperature according to sound-wave propagation data.

Since the velocity of sound waves in a gas increases, approximately, as the square root of the absolute temperature and does not depend on pressure, the above assumption means that somewhere at great heights the air becomes warmer with increasing altitude, and that the sound ray is reflected back toward the earth from a layer of warm air. From systematic measurements of the travel time of sound waves over various distances and of the angles of their arrival, the vertical temperature distribution was obtained11, 12, 13. Some of the results are shown in Fig. 5. Whether these data refer to day or night is not indicated. To date, not a single investigation has been published concerning diurnal variations of temperature. The course of the curve of the vertical temperature distribution above the maximum, located at about 60 km, cannot be obtained from measurements by means of sound rays.

It could have been supposed that the speed of sound increases with height as a consequence of the decrease with height of the mean molecular weight. However, this possibility seems doubtful and finds no confirmation from the standpoint of data testifying in favor of the complete constancy of the composition of the atmosphere up to heights of the order of 100 km.

The theory of tidal oscillations of the atmosphere, developed by Laplace, passed through several interesting phases. The problem consisted in explaining the observed fact that solar semidiurnal barometric variations are approximately 16 times stronger than lunar ones, whereas the tidal force of the Moon is twice as great as that of the Sun. William Thomson in 1882 pointed out that solar pressure oscillations cannot be a thermal effect, since in that case the diurnal variations would be greater than the semidiurnal ones, in direct contradiction with the facts. As a solution to the question, Kelvin proposed the supposition that the atmosphere has a period of its own oscillations close to 12 hours, and that the solar semidiurnal tide is amplified as a result of resonance. Analysis showed that, in order to obtain the necessary amplification, the period of the natural oscillations must be less than 12 hours by 6 minutes. Further evidence concerning oscillations of the atmosphere was obtained from observations of the speed of propagation of the blast wave during the eruption of Krakatau in 1883 and during the fall of the great Siberian meteorite in 1908. In both cases a period of atmospheric oscillations equal to 10.5 hours was found^14.

Fig. 6. Temperature according to the theory of atmospheric tides.

Fig. 6. Temperature according to the theory of atmospheric tides.

The possibility of the existence in the atmosphere of two natural oscillations with periods of 10.5 and 12 hours was studied by Pekeris^15. He found that this property should be possessed by an atmosphere with a vertical temperature distribution of the type of curve 1 in Fig. 6, and that the temperature distributions shown in Fig. 6 by the dashed curves 2, 3, and 4 do not correspond to the presence of natural oscillations with the indicated periods. An argument in favor of the existence of a low temperature at a level of about 80 km, required by curve 1 of Fig. 6, was...

Hemphris’s hypothesis[^16] held that the cirrus-like clouds, visible in rare cases by day at an altitude of 82 km, consist of ice crystals. A subsequent contribution to the problem of atmospheric tides was made by Appleton and Weekes[^17], who detected and measured lunar tides in the E layer of the ionosphere at an altitude of about 110 km, with an amplitude of about 1 km and a maximum approximately $3/4$ hour before the Moon’s passage through the meridian. They came to the conclusion that the force and phase of the tide cannot be reconciled with the above-mentioned theory of atmospheric oscillations. This leaves the theory in an uncertain state.

In order to determine the pressure, and hence also the temperature, of the air at great heights, measurements of the brightness of the sky at the zenith were carried out during twilight[^18]. The method is as follows. The brightness of the sky at the zenith observed during twilight is chiefly the result of the scattering of sunlight in the vertical column of that part of the atmosphere which is illuminated by the direct rays of the Sun. Since the laws of scattering of light by air molecules and the intensity of the direct rays of the Sun are known, from the brightness of the sky at the zenith one can compute the total number of molecules in the sunlit vertical column of air, and hence the pressure at the lower boundary of this column. After sunset, since the angle of depression of the Sun below the horizon increases, the height of the lower boundary of the illuminated region of the atmosphere also increases. Consequently, measurements of the brightness of the sky at the zenith during the progressing evening twilight make it possible to determine the pressures at levels of ever increasing height. A large number of series of such measurements were carried out near Washington at $39^\circ$ north latitude and at Patos (Brazil) at $7^\circ$ south latitude. The results showed that the temperature of the twilight atmosphere is constant with height within $\pm 15^\circ\mathrm{C}$ from 20 to 55 and possibly to 60 km. At these levels the temperature proved to be $220 \pm 15^\circ\mathrm{K}$2.

HIGH LAYERS OF THE EARTH’S ATMOSPHERE

Further, it was found that corrections were necessary for multiple scattering, which was also measured and proved to be small at low altitudes, but to increase rapidly with altitude and, at about 60 km, to become so large that the method becomes completely unsuitable for obtaining information about the atmosphere above 60 km*).

Meteor phenomena were interpreted as indicating that the air density above the level of about 60 km is greater than follows from the assumption of a constant air temperature of 220° K in the interval from 40 to 100 km. It was supposed that the increased air density at these altitudes is due to an increase of temperature above the 40 km level. Thus, meteor observations were used to compute the density of the upper layers of the atmosphere and, from this, their temperature.

The objects of both photographic and visual observations were the following phenomena: the deceleration of meteors and the heights of the beginning, greatest brightness, and end of their visible trajectory. The air density is obtained on the basis of observations of each of these phenomena by means of a specially developed theory, which presupposes known density and sizes of meteors and describes the physical phenomena occurring when a meteor enters the atmosphere, loses energy, and burns up.

A description of the various theories lies outside the scope of my intentions. It is sufficient to present the results summarized in Whipple’s latest paper¹⁹. They are shown in Fig. 7, which coincides with Fig. 2 of Whipple’s paper. The air densities obtained with the aid of the theory from meteor observations are indicated in Fig. 7 by points. These data correspond to nighttime conditions. The solid curve was obtained by calculation on the basis of the temperature distribution with height corresponding to curve 1 of Fig. 6, for which the maximum temperature of 365° K at an altitude of about 60 km falls to 184° K at 80 km. Whipple concludes that “there are considerable fluctuations in the solutions (from meteor data), depending on the weights assigned to the different methods of determining atmospheric density. The best solution appears to be that which corresponds to the solid curve

*) The author’s conclusion as to the unsuitability of the twilight method for studying the atmosphere at altitudes exceeding 60 km, based on his own work¹⁸ cited by him, is erroneous. As a number of studies carried out in the Soviet Union have shown (G. V. Rozenberg, I. A. Khvostikov and F. F. Yudalevich, DAN 58, 1277, 1948; N. M. Shtaude, DAN 59, 1281, 1948; T. G. Megrelishvili and I. A. Khvostikov, DAN 59, 1283, 1948), the role of secondary scattering increases as the Sun sinks below the horizon, but is by no means as catastrophic as Hulburt assumes, and cannot substantially distort the results obtained by the twilight method. In fact, the twilight method is suitable for investigating the upper layers of the atmosphere, in any case up to altitudes of the order of 210, and possibly up to 280–300 km. (Translator’s note.)

Fig. 7. It should be noted, however, that a constant temperature of about $256^\circ$ in the altitude interval from 60 to 100 km does not go too far beyond the range of possible solutions... The data from determinations of the brightness maxima of meteors from photographs of their paths require a zone of high temperature at an altitude of about 60 km, but correspond more to the absence of a zone of low temperature near 82 km...”.

The propagation of sound waves makes it possible to suppose that above the stratosphere the temperature rises approximately to $370^\circ$ K at an altitude of about 60 km. The measurements refer to daytime conditions; for the night there are no measurements. The data do not make it possible to draw conclusions about the temperature above 60 km. The theory of atmospheric tides requires a temperature maximum of about $350^\circ$ K at an altitude of the order of 60 km. Presumably, this is an average value for the equator, but the existence of diurnal variations has not been established. At present the theory encounters the difficulty that it predicts tides in the high layers of the atmosphere which are not consistent with observations over the $E$-layer of the ionosphere. Measurements of the brightness of the twilight sky indicate an air temperature during twilight of $220 \pm 15^\circ$ K in the altitude interval from 20 to 55 km. They give no information on the temperature above 60 km*) or at a time of day other than twilight. Meteor phenomena observed at night indicate an increase of temperature with altitude from approximately 40 to 60 km, but do not provide definite data for judging whether the temperature above 60 km remains constant or decreases.

Fig. 7. Atmospheric density according to meteor-observation data.

Fig. 7. Atmospheric density according to meteor-observation data.

We may note that, although among the data obtained by means of these indirect methods of investigation there are certain discrepancies, evidently none of the methods is so accurate that these discrepancies could be regarded as contradictions**).

) See the note on the preceding page. (Translator’s note.*)

**) The survey given in this paragraph of indirect methods of investigating the high layers of the atmosphere, and also of the results obtained with their aid, …

“EXPERIMENTAL TABLES”

As a result of the preceding consideration, commissions in England, Germany, and America compiled “experimental tables” of the properties of the upper layers of the atmosphere. The temperature curves are shown in Fig. 8; the curve designated NAGA was constructed by the committee of the National Advisory Committee for Aeronautics²). The NAGA curve was called the “experimental standard”; the “experimental maximum” and “experimental minimum” curves, also constructed, are situated respectively 30° above and 80° below the curve shown in Fig. 8. In constructing them it was assumed that both seasonal and diurnal variations of temperature are absent.

Fig. 8. Temperature–height curves based on indirect data; the short curve was obtained from V-2 pressure data of October 10, 1946.

Fig. 8. Temperature–height curves based on indirect data; the short curve was obtained from V-2 pressure data of October 10, 1946.

In the NAGA²) report it was assumed that the structure of the atmosphere remains approximately unchanged from sea level up to 80 km; the mean molecular weight of air was taken as 28.966 times the atomic weight of hydrogen, or equal to $4.816 \cdot 10^{-23}$ g.

With respect to oxygen it was assumed that in the daytime at the level of 80 km it is entirely in the molecular state, and at the level of 100 km entirely in the atomic state, while the concentration of atomic oxygen varies with height according to a linear law. At night the corresponding levels were located at 105 and 120 km. Under these assumptions concerning the temperature and structure of the atmosphere, the “experimental tables” of pressure and density up to an altitude of 120 km were calculated.

[[footnote continued from previous page]] is far from complete. First of all, one should note the generally known data obtained by radiophysical methods and relating to the ionospheric layers $E$ and $F$, the data of Harang, based on the study of the height distribution of the brightness of auroras, and also the data obtained by the twilight method for altitudes greater than 60 km. All these methods lead to sufficiently consistent results, substantially refining and supplementing the picture described by the author, without changing its character, as was shown by T. G. Megrelishvili and I. A. Khvostikov (DAN 59, 1233, 1948...). (Translator’s note.)

E. O. HALBERT

DATA OBTAINED WITH V-2 ROCKETS

In 1945 the Army Artillery Corps received from Germany a considerable number of V-2 rockets and planned the organization of their flights for the purpose of acquiring experience in handling rockets and obtaining data on the upper layers of the atmosphere. The rockets were launched at the White Sands Proving Ground in New Mexico at $33^\circ$ north latitude. To date, data have been published on atmospheric pressure and solar radiation at great heights. Each of these results required great effort and was achieved by specially organized groups of collaborators from the Naval Research Laboratory. All the groups worked within extremely compressed deadlines, since the rocket launches had to take place according to schedule, regardless of whether any particular scientific apparatus was ready. The direct data obtained in a few seconds will, in the end, replace the indirect conclusions of many years of research.

We shall not discuss here the new data that were obtained concerning cosmic rays$^{21}$, since, as was said, cosmic rays have not yet shown themselves to be a significant factor determining the characteristics of the upper layers of the atmosphere.

Pressure measurements$^{22}$ were made during the flight of a V-2 rocket on October 10, 1946, which took place at 11 a.m. 105th meridian time. The readings of the manometers installed on the rocket were transmitted to self-recording apparatus located on the surface of the earth. Two groups of pressure data were obtained. One pertains to the altitude interval from 0 to 12 km and agrees well with the known pressure values at these altitudes. The other group of data covers the altitude interval from approximately 60 to 85 km. The obtained values, shown in Fig. 9 by points, follow the pressure curve constructed by NAGA for a standard day on the basis of indirect data. It is evident from the figure that the values obtained by NAGA are approximately 12% higher than the observed ones.

From the slope of the observed pressure curve (Fig. 9), and under the assumption of such a structure of the upper layers of the atmosphere as was adopted by NAGA, the temperature at the corresponding altitudes was calculated. The values obtained are shown in Fig. 8.

The accuracy of determining the temperature in this way is small, but a critical discussion should be postponed until hopes are realized of obtaining further data with the aid of V-2 rockets*).

) More complete data on pressure and temperature over the entire altitude interval from 0 to 120 km were obtained during the flight of a V-2 rocket on March 7, 1947. (See, for example, the abstract in UFN, XXXIV, issue 3, p. 445.) (Translator’s note.*)

The V-2 rocket, which flew on October 10, 1946, at 11 a.m., was equipped with a spectrograph with a diffraction grating, by means of which, from various altitudes up to an altitude of 88 km, the solar spectrum^23 in the ultraviolet region shorter than 3400 Å was photographed. At 11 a.m. the altitude of the Sun was 51°, solar activity was weak, and the characteristic number of terrestrial magnetism was 0.2 for the period from 0 to 12 hours on October 10. The rocket reached approximately 160 km and fell to earth in the desert of New Mexico at a distance of about 32 km from the launching site. The spectrograph and the cassette with the film were found practically undamaged four days later.

Fig. 9. Pressure data obtained during the flight of the V-2 rocket on October 10, 1946.

After development, 35 spectra were found on the film; the highest was obtained at 88 km. Spectra at greater altitudes were not obtained, since, owing to vibrations of the rocket, the mechanism moving the film operated too rapidly and the entire film was used up on the way to 88 km. Some of the spectra are shown in Fig. 10. All of them were obtained with exposures of 3.6 sec. The altitudes are indicated in kilometers above sea level. Up to 44 km the rocket was stabilized. Above this point it rolled and yawed, as a result of which spectra F and G (Fig. 10) were taken with the Sun in a position displaced from the axis of the spectrograph. Spectrum G, taken at an altitude of 88 km, was exposed so weakly that it shows less ultraviolet than spectrum F, taken at an altitude of 55 km. In Fig. 10, in reproduction, the spectra have been contrasted in order, as far as possible, to bring out the interesting regions.

The spectra reveal an increasing extension into the ultraviolet as the altitude increases. Spectrum D (Fig. 10), taken at an altitude of 25 km, extends to 2925 Å. Spectrum E (34 km) extends to 2650 Å and shows measurable blackening between 2260 and 2100 Å, at the short-wavelength edge of the Hartley ozone absorption band. At the same time, at an altitude of 34 km there still remains, between the spectrograph and the Sun, a sufficient layer of ozone to prevent registration of the spectrum in the central region of the band. Spectrum F, taken—

...taken at 55 km, shows no noticeable absorption by ozone. These spectra of the Sun are the first obtained from great heights above the ozone layer; they extend the region of the solar spectrum known to us in the ultraviolet approximately from 2900 to 2100 Å.

Fig. 10. Solar spectrum obtained from the V-2 rocket on October 10, 1946.

Fig. 10. Solar spectrum obtained from the V-2 rocket on October 10, 1946.

The newly discovered part of the solar spectrum is shown in Fig. 11, on which some absorption lines have been identified.

Fig. 11. Ultraviolet spectrum of the Sun from an altitude of 55 km.

Fig. 11. Ultraviolet spectrum of the Sun from an altitude of 55 km.

This is the same spectrum as spectrum $F$ (Fig. 10), taken at an altitude of 55 km; the reproduction has been contrast-enhanced. The following absorption lines stand out: Si I (2882 Å), strong absorption lines of magnesium Mg I (2852 Å), and the Mg II doublet (2802 and 2795 Å), and the principal lines Fe II (2410, 2405, 2396 and 2382 Å). A detailed analysis of the spectrum has not yet been carried out*).

*) During the flight of the V-2 rocket on March 7, 1947, additional spectra of the Sun from altitudes of 55 and 75 km were obtained. Preliminary data ...

The distribution over the spectrum of the intensity of sunlight that has passed through the entire overlying thickness of the atmosphere was determined by careful photometric comparison of solar spectra obtained at high altitudes with the spectrum of a calibrated carbon arc. Preliminary results are shown in Figs. 12 and 13, in which the new curve in the region of the spectrum shorter than 3000 Å is joined

Fig. 12. Curve of the energy distribution in the ultraviolet spectrum of the Sun.

Fig. 12. Curve of the energy distribution in the ultraviolet spectrum of the Sun.

to the known curve²⁴ for the region of wavelengths greater than 3000 Å, relating to the light of the Sun outside the atmosphere. In Figs. 12 and 13 a curve for a black body is also plotted, chosen so that it passes through the maximum of the intensity curve of sunlight near 4600 Å. It is evident from the figures that the curve relating to the light of the Sun falls off more steeply than the black-body curve and lies below it by factors of about 3 and 10–20 at 3000 and 2200 Å, respectively. This new fact will be important in calculations of the temperature of those atmospheric layers in which solar radiation is absorbed by ozone. Determination of the amount

²⁴ Their processing has been published in Phys. Rev., 71, 827 (1947). A larger number of Fe I and Fe II lines, twenty Si I lines, the C I line (2478 Å), and also lines of the following elements were found: certainly—V I, V II, Cr II, Mn II; probably—Na I, Ni I, Ni II, Cr I, Co II, Be I and Al I; possibly—P I and Cu I. The presence of a number of unresolved absorption bands, probably of molecular origin, is also noted. (Translator’s note.)

of ozone at different heights has not yet been completed. At present it can only be said that on October 10 approximately 3% of all the ozone was located above 34 km and less than 1% above 55 km.

The solar spectra obtained so far provide little information about nitrogen oxides in the high layers of the atmosphere. In the spectra, characteristic bands3 of $NO_2$ or nitrogen tetroxide $N_2O_4$ between 2600 and 2270 Å were not observed. In the case of nitrogen pentoxide $N_2O_5$, Adel and Lampland4 suggested that this gas is responsible for the atmospheric absorption band at $7.6\,\mu$. It is known qualitatively3 that in the ultraviolet this gas has continuous absorption, which increases as the wavelength decreases from 2800 to 2400 Å. But so long as quantitative data are lacking, no information whatever can be obtained from the solar spectrum about the presence or absence of $N_2O_5$ at great heights. This applies equally to nitrous oxide $N_2O$, which has continuous absorption in the region from 3000 to 1760 Å, perhaps too weak to be noticeable in spectra obtained at great heights. Thus, the presence and distribution of various nitrogen oxides in the upper layers of the atmosphere cannot be determined from consideration of spectra obtained at great heights.

Fig. 13. Curve of the energy distribution in the spectrum of the Sun.

LITERATURE

The literature on the high layers of the atmosphere is extensive. The 1523 references in Terrestrial Magnetism and Electricity, Physics of the Earth, series VIII, McGraw-Hill Book Co., I, 1939, cover many aspects of the subject up to 1937. Later references are contained in the 10 reports of the Gassiot Committee of the Royal Society, Reports on Progress in Physics (Phys. Soc., London, 1942–1943), 9, 1–100.

The following works were cited in the present review:

W. J. Humphreys, Physics of the Air (McGraw-Hill Book Co., Inc., New-York, 1940), third edition, fig. 16, p. 44.
L. F. Hafer, Mon. Weather Rev., 68, 125–129 (1940); I. C. Ballard, ibid. 66, 2–9 (1938).
N. K. Johnson, Nature, 157, 24 (1946).
S. Kahlke, Ann. Hydrogr., 49, 294–299 (1921); C. C. Trowbridge, Mon. Weather Rev., 35, 390–397 (1907). Data summarized by E. O. Hulburt, Publ. Astr. Soc. Pacific, 44, 178 (1932).