12.4.1. Brightness Distribution
The three-dimensional structure of extragalactic sources is usually inferred from the observations of the angular size and bright ness distribution projected onto the plane of the sky. Often only the brightness distribution in one dimension is determined. The data on brightness distributions are obtained by one of four common procedures.
a) Pencil Beam, Observations
only a few of the extra-galactic sources are sufficiently large for their structure to be studied by even the largest pencil beam radio telescopes. Moreover, these sources are all relatively nearby, of low absolute luminosity and very low surface brightness, and therefore not representative of typical extragalactic sources found in radio source surveys. One very extended source which has been mapped with pencil beam telescopes is Fornax A (shown in Figure 12.1), which consists of two extended components, with little or no fine structure, located symmetrically on either side of a bright elliptical galaxy.
Figure 12.1 Brightness distribution of Fornax A at 75 cm, mapped with a beamwidth of about 2.'8 of arc beam, using the Molonglo Cross radio telescope in Australia. Contour unit is 100°K. [Taken from Cameron (1969), Proc. Astron. Soc. Australia 1:229].
b) Lunar Occultations
For a number of years the highest resolutions were obtained by observing the diffraction pattern of a radio source as it was occul ted by the Moon. Most of the analysis was based on a technique described by Scheuer (1962) and later extended by von Hoerner (1964) to restore the true brightness distribution from the observed Fresnel diffraction pattern. The maximum resolution is generally limited by the sensitivity of the telescope.
Although a radio-telescope especially designed for this purpose has recently been completed in India, the technique has not enjoyed widespread use for several reasons:
High resolution is obtained only for very strong sources or when using very large antennas, such as the 1000-foot Arecibo spherical reflector. The integration time is limited to about one second by the passage of the Moon across the source, and very short receiver time constants are required to obtain good resolution.
Each occultation gives only a one-dimensional "strip" distribution. Several occultations are required to reconstruct the two-dimensional structure.
At the shorter wavelengths, the Moon is an intense source of thermal radio emission and the small tracking irregularities present in radio-telescopes may completely mask the occultation of the much weaker extra-galactic sources.
Interference from terrestrial radio emission reflected from the Moon is often a serious problem.
c) Interplanetary Scintillations (IPS)
In 1962 a group working at the Cavendish Laboratory in England discovered that radio sources with structure of the order of a second of arc or less showed rapid scintillations when observed through the solar corona. This effect was used extensively for several years to study small-scale-structure radio sources (e.g., Cohen, Gundermann, and Harris, 1967; Little and Hewish, 1966). However, in recent years the high-resolution interferometers now available have been used for this kind of work, and IPS are now rather used to study the interplanetary medium. The reader who is interested in further details about the application of lunar occultations and interplanetary scintillations is referred to the review by Cohen (1969).
d) Interferometry and Aperture Synthesis
During recent years the techniques of interferometry and aperture synthesis have been greatly improved and are now providing accurate and detailed information on the structure of extra-galactic sources. Because of the importance of these techniques, a whole chapter of this book has been devoted to describing the methods in considerable detail.
Most of the sources that have been studied have angular dimensions of less than a few minutes of arc, and about half, of all sources are less than 15 seconds of arc in size. For the resolved sources a simple single-component structure is very rare, and most sources show a surprising amount of structure. Often the source is extended along a single axis, and the most common configuration is the double structure where most of the emission comes from two well-separated components. Frequently the two components are of approximately equal size, and with luminosity as illustrated by the map of Cygnus A shown in Figure 12.2. Typically the overall dimensions are about three times the size of the individual components, which are located symmetrically about the galaxy or QSO. But in some cases the ratio of component separation to component size may be very great. Often an extended low surface brightness component is located between the two primary components, or extends as a tail away from a single bright region, as shown by the radio photograph of the radio galaxies 3C 129 and 3C 129.1 illustrated in Figure 12.3.
Figure 12.2 Brightness distribution of the intense radio galaxy Cygnus A at 6 cm observed with the 6-second-of-arc beam of the Cambridge 1-mile aperture synthesis radio-telescope. [Taken from Mitton and Ryle (1969), Monthly Notice Roy. Astron. Soc. 146:221].
Other sources have been observed to show more complex structures containing three or more widely separated components which may also be aligned along a single axis (Macdonald et al., 1968; Fomalont, 1969). In all of these extended sources the radio emission comes primarily from regions well removed from the optical galaxy or quasar. High-resolution studies of these extended clouds often show considerable structure, with one or more highly condensed regions existing many kiloparsecs from what is presumed to be the parent galaxy or quasi-stellar object. Typical linear dimensions for these extended sources range from about 10 kpc to several hundred kiloparsecs.
Following the early detection of apparent extensive halos around two nearby galaxies, M31 and M33, and the speculation of our own Galactic halo (Chapter 1), it was widely supposed that the radio emission from "normal" galaxies originated in a much larger volume of space than the optical emission. This is not the case, however, and in fact in general the "normal" galaxy emission is often highly concentrated to the galactic nucleus.
In many objects, such as quasars or Seyfert and N galaxies, but also including some elliptical galaxies of normal appearance, there are one or more very compact radio sources coincident with the region of brightest optical luminosity. These compact sources also have complex structures with component sizes ranging from about 0."1 arc to well under 0."001 are. Often compact and extended components exist simultaneously. The extended components may appear as (1) a "halo" surrounding a compact "core" component, as in the galaxy M87, (2) a jet extending away from the compact component as in 3C 273, or (3) a pair of unconnected extended components lying on either side of the compact central sources, as in 3C 111. Generally the sources in categories (1) and (3) are identified with E galaxies and those in (2) with N or Seyfert galaxies, or quasars.
The observations of the sub-structure of the compact sources are made with independent oscillator tape-recording interferometers using telescopes widely spaced around the world with base-lines up to 80% of the Earth's diameter. Since intermediate baselines are insufficiently sampled, the data are not adequate to completely reconstruct the brightness distribution and it is necessary to resort to model fitting. To the extent that the brightness distribution can be inferred from the limited data, the structure of the compact sources appears remarkably similar to that of the extended sources, in the sense that in general they do not show circular symmetry, but consist of two or more well-separated components, lying along a single axis. Thus over a range of angular (and linear) dimension of about a factor of 105, the radio sources show essentially similar structure-only the scale size varies.
The smallest linear size which has been directly measured is the compact source located in the nucleus of the nearby galaxy M87, which contains about 1% of the total flux density and is only about 0.1 pc across in extent.
One of the best-studied sources is the intense radio galaxy Cygnus A, shown in Figure 12.2, which contains two major com ponents separated by about 2 minutes of arc, with a galaxy located halfway between. Each main component is about 20 seconds of arc in size, and contains a faint tail somewhat elongated along the line joining the components. Near the outer edge of each of the main components is an intense bright core. High-resolution observations of the western core show that the core itself contains a double source with a separation of about 5 arc seconds along a position angle about 20° from the line joining the two main components. The eastern component is also a double with a separation of a few arc seconds along a line nearly perpendicular to the line joining the main components. Each of the subcomponents has an angular size of the order of one arc second (Miley and Wade, 1971). Also there is a weak, compact component coincident with the optically identified galaxy.
Some sources, particularly those located in dense clusters, have an extended region of low surface brightness emission. It has been postulated that these are trails of relativistic electrons left by the motion of a radio galaxy through a gaseous medium. The radio photograph of 3C 129 and 3C 129.1 shown in Figure 12.3 clearly shows a double-helix trail suggesting rotation of the particle-ejecting region of the galaxy (Miley et al., 1972).