DETERMINATION OF THE STRUCTURE OF THE OZONE LAYER UP TO AN ALTITUDE OF 70 km

Indirect methods for determining the structure of the ozone layer (for example, by means of the so-called Umkehr effect) have made it possible to obtain fairly reliable data on the ozone content only up to altitudes of about 50 km1. This boundary was not only not surpassed, but was not even reached as a result of measurements made by raising spectrographs on rockets2. Meanwhile, obtaining detailed information on the structure of the upper boundary of the ozone layer appears to be very important both from the point of view of refining the photochemical theory of layer formation and of explaining the temperature maximum at altitudes of the order of 50 km, and, in general, for the development of our ideas about the formation and structure of ionospheric layers, the mechanism of night-sky glow, and other processes occurring in the high layers of the atmosphere. Let us recall that, because of the absence of the necessary data on the cross sections of various kinds of reactions and collisions that may occur under stratospheric conditions, estimates of the ozone content in the upper part of the layer, made on the basis of photochemical theory, differ by as much as a factor of 1000. Therefore, the direct determination of the ozone content at altitudes up to 70 km, carried out by the authors of the paper under review3, is of undoubted interest.

In its main features, the measurement procedure was identical to that used by the same authors earlier2. With the aid of an automatic spectrograph lifted on a rocket, a set of spectra of direct sunlight was obtained, taken from different depths of the ozone layer. However, along with the spectrograph that had been used in the previous work (spectrograph A) and whose description may be found, for example, in4, a spectrograph of another design (B), shown in Fig. 1 at the top, was also used. In essence, it was a combination, in one housing, of two independent spectrographs with aluminum gratings (radius of curvature 40 cm, 15,000 lines per inch). The main difference lay in the arrangement of the light-guiding system, which was intended to ensure that direct rays of the Sun entered the spectrograph under conditions of rotation and yawing of the rocket.

Instead of the fluorine-lithium sphere that had replaced the slit in the spectrograph of the previous type, here a slit of a special design was used, formed by a pair of mirror plates inclined toward one another (Fig. 1, bottom). The slit proper was a gap (0.03 mm wide and 2 mm long) at the apex of the angle formed by the mirrors. The angle between the mirrors was 10°. The divergence of the beam illuminating the grating was 5°. As a result, rays falling on the slit in the following interval of angles with respect to the bisector of the system of mirrors reached the grating: \(0^\circ \pm 2.5^\circ\)—direct rays; \(\pm(10^\circ \pm 2.5^\circ)\)—after a single reflection from a mirror; \(\pm(20^\circ \pm 2.5^\circ)\)—after a double reflection; and \(\pm(30^\circ \pm 2.5^\circ)\)—after a triple reflection. The bisectors of the slits corresponding to the two halves of the spectrograph were inclined relative to one another by an angle of 5° (see Fig. 1, top), as a result of which the field of view of the spectrograph contained (allowing for the rotation of the rocket) an interval of angles of \(\pm 30^\circ\) with respect to the horizon.

Because there were no other light-directing devices, the field of view of each half of the spectrograph was limited in the horizontal plane to an angle of 4°. The film was in continuous motion at a speed of 2 mm/sec. The rate of rotation of the rocket about its axis (about 1 revolution per second) limited the exposure time. (In the spectrograph of the old design the exposure time was 1 sec, with an interval of 0.2 sec.) For the quantitative measurement

of the film’s spectral intensity distribution was preliminarily calibrated by means of a standardized carbon arc and a sector disk.

The launch was carried out on June 14, 1949, at White Sands (32°24.4′ N latitude and 106°20.4′ W longitude) at 19 hours 03 minutes local time.

Having reached an altitude of 112 km, the rocket fell 34.77 miles north and 3.4 miles west of the launch site. During the flight, precise trajectory data were obtained, which made it possible to calculate the longitude and altitude of the rocket at the moment of exposure of each of the spectra. In all, more than 200 spectra were obtained, covering the altitude interval from 19 to 110 km; however, some of them proved unsuitable for further processing.

Fig. 1. Above—the layout of the spectrograph; below—the layout of the slit.

A distinctive feature of these measurements was that, at the moment of the rocket’s launch, the Sun was very near the horizon (about 1°), as a result of which the path of the rays through the atmosphere was considerably lengthened, and the absorption of light by atmospheric ozone, very weak above 50 km, proved measurable. However, even under these conditions, above 70 km the absorption became so insignificant that the spectra obtained at these altitudes practically did not differ from one another, and determination of the ozone concentration became impossible. The authors note that in this altitude interval (70–110 km) no other gases were detected whose absorption was noticeably less than that of ozone.

Determination of absorption by ozone was carried out by photometric comparison of spectra obtained at different altitudes with spectra obtained at altitudes > 70 km. In doing so, it was not limited to determining the intensity at each wavelength, but was carried out in a sufficiently broad spectral interval, which substantially reduced the measurement error.

For the altitude interval 60–70 km the entire region of the ozone absorption band was used, down to 2500 Å. After allowance for Rayleigh scattering and the introduction of other corrections, the authors obtained, for the amount of ozone along the path of a light ray as a function of the rocket altitude, the data shown in Fig. 2 (the amount of ozone, as usual, is indicated in units of the thickness of an equivalent layer of pure ozone under normal conditions). It is seen from the figure that the data obtained with different spectrographs are in good agreement with one another and that the measurement error is small.

Fig. 2. Measurement results (ozone thickness in the direction toward the Sun).

Since the Sun was near the horizon, it was necessary to take into account the curvature of the Earth’s surface (and, correspondingly, of the ozone layers). To do this the authors divided the atmosphere into spherical layers 2 km thick and calculated the path of the ray through each of the layers as a function of the rocket altitude. Assuming homogeneity of the layers throughout their extent, for the amount $\tau_i$ of ozone located along the ray path up to the $i$-th level, the following approximate relation is obtained

Fig. 3. Ozone concentration as a function of altitude.

\[ \tau_i = \sum_j \rho_i\,(\Delta s_i)_j, \]

where $\rho_j$ is the ozone density in the $j$-th layer, and $(\Delta s_i)_j$ is the path of the ray through the $j$-th layer when the observation is made from the $i$-th layer.

Further, the authors linearly extrapolated the curve of Fig. 2 to a height of 87 km (where the practically complete absence of ozone was assumed), and then

successively found the values \(\rho_j\) on the basis of the relations:

\[ \tau_{85}=\rho_{86}(\Delta s_{85})_{86}; \]

\[ \tau_{83}=\rho_{84}(\Delta s_{83})_{84}+\rho_{86}(\Delta s_{83})_{86}\quad \text{and so on.} \]

The authors note that the effect of extrapolating the curve of Fig. 2 above the level of 70 km is appreciable only above 65 km (at 70 km the neglect of absorption by the overlying layers gives an error of 100%).

The results obtained are given in the table and in Fig. 3. Fig. 3 also presents data obtained during previous ascents.\(^{2}\) The discrepancies are quite natural, since the form and position of the maximum must depend on solar activity and meteorological conditions. The agreement with data obtained by indirect methods (the Umkehr effect) is satisfactory. In Fig. 4 the same data are plotted on a logarithmic scale. As is evident from the figure, above 35 km the ozone concentration varies with height practically according to an exponential law.

Fig. 4. Ozone concentration as a function of height.

Figure 5 shows the dependence of the total amount of ozone located above a given level on altitude. Measurements made on the same day at the same geographical point from the earth’s surface gave a value \(\tau = 1.9\) mm, and at the mountain altitude (altitude 9200 feet, distance from the rocket-launching site 40 miles) \(\tau = 1.78 \div 1.86\) mm; both values fit well on the extrapolation (dashed) curve for low altitudes.

Fig. 5. Total amount of ozone in a vertical column of air above a given level. The crosses show the results of measurements from the earth’s surface.

In conclusion, the authors compare the results they obtained with various versions of the photochemical theory (including their own version) and arrive at the conclusion that the experimentally measured change in ozone concentration with altitude above 40 km is in good agreement with the photochemical theory, but only if triple collisions are taken into account.

T. R.

References Cited

1. See, for example, F. W. Götz, A. R. Meetham and G. M. B. Dobson, Proc. Roy. Soc., 145 A, 416 (1934); 148, 598 (1935); E. Tonsberg and K. L. Olsen, Geofys. Pub., 13, No. 12 (1944); R. V. Karandikar and K. R. Ramanathan, Proc. Ind. Acad. Sci. 29 A, 330 (1949).

2. F. S. Johnson, J. D. Purcell and R. Tousey, J. Geophys. Res., 56, 383 (1951).

3. F. S. Johnson, J. D. Purcell, R. Tousey and K. Watanabe, J. Geophys. Res. 57, No. 2, 157 (1952).

4. S. L. Mandel’shtam, UFN, 46, 145 (1952); G. V. Rozenberg, UFN, 31, 281 (1947).