From Current Literature

RADIO ASTRONOMICAL METHOD FOR STUDYING THE UPPER LAYERS OF THE ATMOSPHERE

Observation of extraterrestrial radio emission, besides its direct interest in connection with the study of its sources (the Sun, “radio stars,” etc.), also makes it possible to investigate the Earth’s atmosphere, and in particular to obtain information about ionospheric layers needed for satisfactory calculation of the passage of radio waves on communication lines. Since here we are dealing with a new method, it is interesting to consider to what extent radio-astronomical data agree with data obtained with the aid of ordinary ionospheric apparatus, and what information, inaccessible to ordinary methods, can be extracted from them.

At present there is very little information about layers lying above the layer \(F_2\). Ordinary ionospheric technique cannot provide such data because possible reflections from these layers are completely screened by the layer \(F_2\), which possesses the greatest critical frequency. Indications of the existence of ionization above the layer \(F_2\) have been obtained in the form of so-called “spread” reflections from the layer \(F_2\) and reflections from the layer \(G\). The former are believed to be caused by inhomogeneities with increased ionization density, located above the level of maximum ionization of the layer \(F_2\), while the latter by a regular layer observed in rare cases above the layer \(F_2\), when the density of ionization of the latter falls below the density of the layer \(G\). However, such a conclusion is not sufficiently reliable and convincing, for there are many objections to the assumption of the existence of the layer \(G\).

To solve the question of the existence and character of ionization above the layer \(F_2\), radio-astronomical methods may be brought to bear, since extraterrestrial radio waves, passing through the entire thickness of the Earth’s atmosphere and undergoing its influence, will bear the imprint also of passage through ionized layers. Thus, they can serve as an additional source of information about the upper layers of the atmosphere.

During radio-astronomical observations the following ionospheric phenomena have already been studied:

1) ionospheric refraction,
2) absorption in the ionosphere,
3) “scintillation of radio stars,”
4) sudden ionospheric disturbances.

Refraction and absorption in the ionosphere.
The position of a radio-emission source on the celestial sphere is determined by the apparent angle of declination, measured on the Earth with the aid of a radio telescope. This angle is the sum of the angle of declination of the source without taking account of the influence of the atmosphere and the angle of refraction undergone by the radio wave from the source in passing through the ionosphere and troposphere. When ionospheric refraction is determined, it is assumed that the tropospheric refraction

and the physical position of the source do not depend on frequency and time. Therefore, if the angles of inclination of a source are measured, for example, at two different frequencies simultaneously, then the difference of the measured angles determines only the difference of the refraction angles in the ionosphere at these frequencies. It is obvious that the angle of refraction in the ionosphere depends on how much the frequency of the received radiation from an extraterrestrial source differs from the critical frequency of the \(F_2\) layer. At frequencies much higher than the critical one, refraction will be small, since the radio wave penetrates the ionosphere almost without deviating from a rectilinear path; at frequencies close to the critical one, the angle of refraction will be greatest; and, finally, extraterrestrial radiation at frequencies lower than the critical one does not reach the Earth at all, being completely reflected from the \(F_2\) layer.

Measurements of the refraction of solar noise were made simultaneously at two frequencies: 60 and 200 Mc/s. In radiolocating the Moon at a frequency of 20 Mc/s, considerable angles of ionospheric refraction were also observed.^3 In both cases, in order to compare the results obtained with theory, the values of refraction in the ionosphere were calculated from the known formulas for a flat Earth and a parabolic distribution of ionization density in the upper layers of the atmosphere; corrections for the curvature of the Earth’s surface were also computed. It turned out that the measured angles of ionospheric refraction were approximately three times greater than the calculated ones.

In the study of ionospheric absorption,^3 considerable discrepancies with theory were also established. These facts served as the basis for a whole series of hypotheses proposed in order to reconcile the observed discrepancies between measurements and theory. Such are the hypotheses of the existence of a \(G\) layer, of the presence^3 of inhomogeneities of considerable size in the \(F_2\) layer, of a noticeable difference between the distribution of ionization in the \(F_2\) layer and the parabolic law^1, and others. However, so far the existing hypotheses have not yet been sufficiently exhaustive in substantiating the assumptions put forward. Therefore, even now there are no final conclusions.

“Twinkling of radio stars” (motion of inhomogeneities). The phenomenon of the “twinkling of radio stars,” or of discrete sources of extraterrestrial radio waves, consists in the fact that the received noise power fluctuates continuously, having the character of random fadings.

In this case the mean period of the fluctuations is about one minute, while the mean power of the source remains constant. Extraterrestrial radiation is received on antennas with high directivity and large gain factors. The received radiation is then amplified by a receiver possessing a low intrinsic noise level and a stabilized gain factor. At the receiver output an ordinary loop oscillograph is usually connected, which records on a slowly moving tape the fluctuations, received by the antenna and amplified by the receiver, of the noise of a discrete source. As the latter, the most powerful of the known sources of extraterrestrial radio waves are almost always used, namely, the sources in the constellation Cassiopeia (declination \(58^\circ 32'\), right ascension \(23^\mathrm{h}21^\mathrm{m}\)) and in the constellation Cygnus (declination \(40^\circ 35'\), right ascension \(19^\mathrm{h}57^\mathrm{m}\)). At a wavelength of \(3.7\ \mathrm{m}\) they have a radiation intensity^5 of the order of \(2 \cdot 10^{-22}\ \mathrm{W}\ \mathrm{m}^{-2}\ \mathrm{cps}^{-1}\). Proof of the terrestrial origin of the fluctuations was obtained by observations on separated receivers.^4 The receivers, together with their antennas (radio telescopes), were installed at some distance from one another. The presence of correlation between the fluctuations at these points was determined. It turned out that the correlation begins to disappear when the receivers are separated by a distance of more than four kilometers. This indicates that the cause of the fadings lies not somewhere in the interstellar medium, but in the Earth’s atmosphere. Indeed—

indeed, if this were not so, then with a freezing duration of about 30 sec the dimensions of the diffraction pattern would have to be of the order of 900 km (the Earth’s orbital velocity is \(30\ \mathrm{km\ sec}^{-1}\)). The absence of correlation between records obtained at receiving points spaced apart by a distance considerably less than 900 km shows that the actually observed diffraction pattern has much smaller dimensions. Consequently, the results of the observations can be explained only on the assumption that the diffraction pattern moves together with the Earth and is of terrestrial origin. The source of the fluctuations in connection with this is naturally to be regarded as the ionosphere, or else it is subject to irregular changes capable of noticeably affecting the propagation of radio waves. Comparison of observations of scintillations over a considerable period of time with strong disturbances of the \(E\) layer do not correlate with one another. Study of the correlation with disturbances of the higher and more strongly ionized \(F\) layer gave comparatively good results. A close connection is observed between the fluctuations and the “stretched,” or diffuse, reflections from the \(F_2\) layer. Both the former and the latter show a rapid increase from 20 h 00 m to 22 h 00 m, have a maximum at 01 h 00 m local time, and then decrease toward morning. At present it is accepted that fadings on meter waves of extraterrestrial radiation are the result of the passage of the latter through an irregular diffraction screen consisting of electron clouds in the \(F_2\) layer of the ionosphere. The frequencies at which fadings are observed are high, and therefore absorption in the ionosphere will be very small and is not taken into account in calculations. Inhomogeneities of ionization in this case will cause only irregular phase changes in the passing wave front, just as occurs in the case of light passing through a plate of nonuniform thickness (this question is discussed in detail in the collection Problems of Modern Physics, 1953, issue 11, pp. 171—176). The height of this diffraction screen is not known precisely, but it is believed to lie between the maximum of ionization in the \(F_2\) layer and the exosphere.

Radioastronomical observations carried out simultaneously on three receivers installed at the vertices of an equilateral triangle with side length 2—3 km made it possible to determine fairly small differences in the times of appearance of individual fluctuations at the corresponding stations \(^{6}\). From the differences in the times of appearance of the fluctuations, the velocity and direction of motion of the diffraction pattern along the Earth’s surface were calculated, which are determined by the drift of electron clouds in the ionosphere. As a result of numerous observations \(^{5}\) it was established that the mean drift velocity of the inhomogeneities causing the fluctuations is equal to \(200\ \mathrm{m\ sec}^{-1}\). The velocities from one night to another, in general, vary considerably and show a close connection with disturbances of the Earth’s magnetic field. They are constant under stable geomagnetic conditions, for example, during any given night. The motion of the diffraction screen is almost always directed along the east—west line, with the predominant direction before midnight being westward, and after midnight eastward. The change in the direction of motion at midnight occurs in a very short time (30—60 minutes) and is characterized by appreciable constancy.

A linear relation was noted between the drift velocity and the frequency of fluctuations (the number of maxima per minute); moreover, for velocities in the range \(30—1000\ \mathrm{m\ sec}^{-1}\) this relation has a scatter of only 20%. This circumstance made it possible to compile a table for determining the drift velocities of the diffraction screen from the fluctuation velocity, which greatly facilitated the processing of observational records (see Table 1).

It was further suggested that, just as the change in the frequency of scintillation of visible stars is connected with changes in wind speed in the troposphere, the change in the frequency of fluctuations of radio noise from radio stars is caused

changes in the velocity of motion of inhomogeneities in the upper layers of the ionosphere. Such an assumption seems plausible in the light of the data presented above (Table I).

Table I

Recently it has become evident that the entire region of the atmosphere, from heights of the order of 90 km up to the very highest parts of the \(F_2\) layer, is subject to winds of high velocity. Evidence for the existence of winds has been obtained from sources of various kinds, for example, observations of meteor trails, silvery clouds, luminous bands, measurements of moving inhomogeneities, and others. For comparison, the results are summarized in Table II²:

Table II

Simultaneous observations of fluctuations of radio stars that have different coordinates on the celestial sphere make it possible to observe and compare motions in the ionosphere at sites lying some distance apart from one another. Thus, for example, circum-polar sources in the constellations Cygnus and Cassiopeia, differing in right ascension by \(3^h 24^m\), successively pass through their upper and lower culminations, during which they can be registered by means of radio telescopes. Since the angles of declination of the sources are different, regions differing in geographical position are observed in the ionosphere as well. Thus it was possible to show\({}^5\) that in regions of the ionosphere separated by distances of about 800 km from one another, the directions of drift coincide within 10°, while the difference in drift velocity is less than 30%. Since the observed regions, in their geographical position, correspond to different latitudes, the supposition was made that zones having a considerable extent in the latitudinal direction are embraced by uniform winds of large scales. Of course, more detailed observations than those already carried out are needed to establish the exact degree of correlation between the velocities and directions of motion of irregularities in the \(F_2\) layer at large distances between the observing points.

Very interesting are the results of observations carried out from middle latitudes (England) of the frequency of fluctuations in the zone of polar auroras. It is known that at high latitudes the ionosphere is almost always in a disturbed state. This is confirmed by the increased frequency of fluctuations from this region. The measured mean velocity of motion of irregularities here is \(360\ \text{m sec}^{-1}\), instead of \(200\ \text{m sec}^{-1}\) in temperate latitudes.

Table III

* The question mark denotes investigations not yet carried out.

At a wavelength of 3.7 m, fadings of extraterrestrial radiation were recorded during auroras and magnetic storms.^6 A complete correlation was noted between the frequency of the fluctuations and the intensity of the phenomena named. The frequency of the fluctuations in their presence increased by approximately a factor of 4. The investigations showed that the increase in the frequency of fluctuations during auroras and magnetic storms is associated with an increase in the velocity of ionospheric winds and that, in general, the velocity of these winds is proportional to the geomagnetic index $K$.

These facts have found very interesting confirmations in the theory of geomagnetic storms and diurnal disturbances in the atmosphere.^7 According to this theory, magnetic disturbances occur because of electric currents circulating in the ionosphere, and these latter are caused by potentials arising as a result of charge separation of moving ionized particles in the presence of the Earth’s magnetic field. In temperate latitudes the diurnal component of the magnetic disturbance requires zonal winds on a large scale, which blow westward from 12:00 to 00:00 local time and eastward thereafter.^7

Investigations of the atmosphere by the radio-astronomical method have only just begun and are still far from completion. Possible directions for such investigations are presented in Table III^1 (see p. 457).

N. T.

References Cited

1. A. P. Mitra, Ind. J. Phys. 26, 495 (1952).
2. R. Payne-Scott, L. McCready, Ter. Mag. Atmos. Elec. 53, 429 (1948).
3. F. I. Kerr, C. A. Shain, PIRE 39, 230 (1951).
4. C. Little, A. Maxwell, Philos. Mag. 42, 267 (1951).
5. A. Maxwell, Dagg. Philos. Mag. 45, 551 (1954).
6. C. Little, A. Maxwell, J. Atm. and Ter. Phys. 2, 356 (1952).
7. E. H. Vestine, J. Geophys. Res. 58, 539 (1953).