THE ULTRAVIOLET SPECTRUM OF THE SUN, OBTAINED FROM ALTITUDES UP TO 88 KILOMETERS

For an observer located on the Earth’s surface, the spectrum of the Sun, as well as the spectra of other extraterrestrial light sources, is sharply cut off at the short-wavelength end near 2900 Å. The cause of this cutoff in the spectrum is, as is well known, the selective absorption of light by atmospheric ozone. The principal absorption band of ozone (the Hartley band) lies in the interval of approximately 2100–3300 Å, with a maximum near 2550 Å. In the region 2300–2800 Å the absorption coefficient is so large that an ozone layer of 0.1 mm at atmospheric pressure is sufficient for practically complete absorption of all solar radiation. The total ozone content in the atmosphere corresponds to a layer about 3 mm thick at 760 mm Hg.

Beginning at approximately 2100 Å and shorter wavelengths, oxygen is chiefly responsible for the absorption, having in this region a broad and very intense absorption band. At the boundary of the absorption bands of oxygen and ozone—in the interval 2100–2200 Å—the transparency of the atmosphere is somewhat higher, and in 1934 Göetz, Schein, and Stoll succeeded, with the aid of a Geiger–Müller counter, in detecting traces of solar radiation in this region of the spectrum.

Since ozone absorbs all the energy of the Sun’s radiation in the far ultraviolet, as well as a considerable fraction of the infrared radiation at the Earth’s surface, it has a substantial influence on the thermics and dynamics of the atmosphere, and knowledge of the distribution of ozone with height is of exceptional importance for atmospheric physics. There is no need to speak of the significance that knowledge of the ultraviolet spectrum of the Sun would have for astrophysics.

The importance of these problems has given rise to numerous attempts, on the one hand, to penetrate, at least partially, into the region close to 2900 Å and into the region of the “window” near 2100 Å through the ozone “roof,” and, on the other hand, to determine the structure of the ozone layer both by directly raising apparatus to great heights (Regener in 1931 succeeded in photographing the spectrum of the Sun from an altitude of 31 km, having raised a spectrograph on a balloon-sonde) and by indirect methods.

It has been established that the principal mass of ozone is concentrated in a layer extending in height to about 55 km, with its center of gravity at about 25 km above sea level. Consequently, the complete ultraviolet spectrum of the Sun in the region of the ozone absorption band could be expected to be obtained only by raising a spectrograph to altitudes of at least 55 km. Until recently such altitudes were inaccessible.

On October 10, 1946, during a test flight of a V-2 rocket projectile at White Sands Proving Grounds (New Mexico, U.S.A.), which reached an altitude of 160 km, a series of spectra of the Sun was obtained from altitudes up to 88 km*).

The spectrograph on which these spectra were obtained consisted of a concave aluminum diffraction grating of 15,000 lines per inch, with a radius of curvature of 40 cm and an aperture ratio of \(f:10\). The dispersion was 44 Å per mm. The resolving—

*) W. A. Baum, F. S. Jonson, J. J. Oberly, C. C. Rockwood, C. V. Strain and R. Tonsey. Phys. Rev. 70, No. 9–10, 781 (1946).

thrust—3 Å. Instead of a target, a spherical bead of lithium fluoride 2 mm in diameter was used, apparently mounted on the side surface of the rocket. It produced a small real image of the solar disk, acting as a light source and, owing to the astigmatism of the grating, producing spectral lines. The authors indicate that this system has a very wide field of view and is much more effective than a target of equivalent width covered with a matte plate. On the opposite wall of the rocket there was a second bead. Plane mirrors on each side directed both light beams onto the grating. Owing to this arrangement, the probability of direct rays of the Sun entering the spectrograph during the rotation of the rocket was doubled.

The spectra were photographed on film sensitized for the ultraviolet; an eight-second cycle of three independent exposures of 0.12, 0.66, and 3.6 seconds was provided. Of the one hundred possible photographs, for technical reasons a series was obtained consisting of only 35 spectra taken at different altitudes. Above 88 km the photography was unsuccessful, mainly because of vibration of the rocket.

Since the air pressure in the spectrograph coincided with the external atmospheric pressure at the corresponding altitude, the short-wavelength limit of the spectrum was determined only by the transparency of lithium fluoride (about 1100 Å). The long-wavelength limit was determined by the design of the instrument and was at 3400 Å.

The figure presents spectra taken at different altitudes with an exposure of 3.6 seconds. At the left are indicated the altitudes (in kilometers) from which the spectra were obtained. The wavelengths are given in angstroms. The contrast of the spectra in the reproduction has been reduced in order, as far as possible, to bring out the region of interest.

At about 44 km the rocket stabilized. Above this it spun and “yawed,” as a result of which spectra \(F\) and \(G\) were obtained under conditions in which the Sun was favorably positioned relative to the axis of the instrument. Consequently, the relative weakness of the ultraviolet portion in spectrum \(G\) (now the spectrum obtained from the greatest altitude) in comparison with spectrum \(F\) should be explained by the shortness of the actual exposure. The blurring of spectra \(E\) and \(F\) was caused by vibration and rotation of the rocket. The fact that the spectra weaken, especially along the lines, during rotation of the rocket is characteristic of the bead-slit device used. In spectra \(F\) and \(G\) this effect appeared in the displacement of the spectra along the lines during the exposure.

From examination of the spectra it is evident that, as the altitude increases, they are progressively extended toward the short-wavelength side. At 25 km the spectrum

photographed down to 2925 Å. Spectrum E (34 km) already extends to 2650 Å and, in addition, shows a weak region approximately from 2100 to 2260 Å, not visible in the reproduction. Above 34 km there still remains above the instrument a sufficient amount of ozone (according to Regener’s data, less than 30%) to prevent the appearance of the spectrum in the center of the Hartley band, but the spectrum in the “window” between the oxygen and ozone absorption bands is already visible. At 55 km the principal mass of ozone has already been passed through, and the insignificant amount of it remaining above the instrument, although still clearly perceptible, already permits easy photography of the solar spectrum over the entire region of the Hartley band. As was to be expected, it proved impossible to obtain the spectrum of the Sun in the Schumann region, which is important both for estimating the amount of oxygen above 55 km and for developing ideas about the mechanism of formation of the ozone layer. Of undoubted interest from the geophysical point of view are also several unexpected absorption bands in the region of 2800 Å, belonging neither to ozone nor to oxygen. The processing of the spectra has not yet been completed and, apart from their reproduction, no data have been published.

It should be noted that these spectra were obtained with apparatus of very low light-gathering power and with comparatively short exposures.

Apparently, on the same V-2, for the same purposes, another spectrograph was also carried aloft, specially constructed at Harvard University by Van Allen and Hopfield. Photographs of it may be found in Popular Science (October 1946, p. 78), but whether any results have been obtained with its aid is as yet unknown.

G. Rosenberg