An Introduction to Active Galactic Nuclei

1.3.1 Radio Properties of Quasars

The radio morphology of quasars and radio galaxies is often described broadly in terms of two components, ``extended'' (i.e., spatially resolved) and ``compact'' (i.e., unresolved at ~ 1" resolution), that have different spectral characteristics, although the synchrotron mechanism seems to be at work in both cases. The extended-component morphology is generally double, i.e., with two `lobes' of radio emission more or less symmetrically located on either side of the optical quasar or center of the galaxy. The linear extent of the extended sources can be as large as megaparsecs. The position of the optical quasar is often coincident with that of a compact radio source. The major difference between the extended and compact components is that the extended component is optically thin to its own radio-energy synchrotron emission, whereas this is not true for the compact sources.

Although a detailed discussion of synchrotron radiation is beyond the scope of this book, we will summarize some of the basic properties of synchrotron-emitting sources. For a homogeneous source with constant magnetic field B, a power-law continuum spectrum (eq. 1.5) can be generated by the synchrotron mechanism by an initial power-law distribution of electron energies E of the form

Equation 8 (1.8)

where alpha = (s - 1) / 2. For the extended component, a typical observed value of the power-law index is < alpha > approx 0.7, so < s > approx 2.4. This applies at higher frequencies where synchrotron self-absorption is not important. At lower frequencies, the emitting gas is optically thick and the the trend towards increasing flux with decreasing frequency turns over to yield F propto ^5/2. The turn-over frequency increases with the density of relativistic electrons in the source, although it depends on other parameters as well. The relativistic particle densities in the extended radio components are low enough that they are optically thin even at very low radio frequencies (at least to the 3C frequency of 158 MHz). Radio spectra sometimes curve downward at higher frequencies (i.e., alpha increases with ). The basic reason for this is that electrons radiate at frequencies proportional to their energies E, and the rate at which they lose energy is proportional to E². Thus, the highest-energy electrons radiate away their energy the most rapidly, thus depleting the emitted spectrum at the high-frequency end first if no replenishment of the high-energy electrons occurs.

Extended radio structures can be divided into two separate luminosity classes (Fanaroff and Riley 1974). Class I (FR I) sources are weaker radio sources which are brightest in the center, with decreasing surface brightness towards the edges. In contrast, the more luminous FR II sources are limb-brightened, and often show regions of enhanced emission either at the edge of the radio structure or embedded within the structure. Bridle and Perley (1984) give L (1.4 GHz) = 10³² ergs s^-1 Hz^-1 as the transition specific luminosity between the two types. Quasars are FR II sources. Examples of FR I and FR II sources are shown in Figs. 1.4 and 1.5, respectively.

Figure 1.4. Example of an Fanaroff-Riley type I (FR I) galaxy. This is a radio map of the Virgo cluster elliptical galaxy M84 (3C 272.1 = NGC 4373), based on 4.9 GHz data obtained with the VLA by Laing and Bridle (1987). Data courtesy of A. H. Bridle, figure by R. W. Pogge.

The characteristics of compact sources are quite different from those of the extended sources. Very long-baseline interferometry (VLBI) yields upper limits on the sizes of compact sources typically no better than ~ 0.01 pc. In contrast to the steep spectra of extended sources, the radio spectra of compact sources are usually flat, i.e., with 0.5, if a power-law form is assumed. Unless the initial electron energy distribution in the compact source is quite flat and quite different from that which drives the extended sources, a synchrotron spectrum is only approximately flat near the turnover frequency and even then only over a limited frequency range. The flatness of compact source spectra over several orders of magnitude in frequency is thus usually attributed to either source inhomogeneity or the presence of a number of unresolved small discrete sources within the compact core. In either case, different parts of the compact region become optically thick at different frequencies, which can flatten the integrated spectrum of the compact source over a suitably broad frequency range.

Figure 1.5. Example of an Fanaroff-Riley type II (FR II) galaxy. This is a map of the z = 0.768 quasar 3C 175, as observed with the VLA at 4.9 GHz (Bridle et al. 1994). The quasar itself is coincident with the bright compact source at ( Delta alpha , Delta delta ) = (0, 0). A jet extending from the compact source to the extended radio lobes is observed on one side of the source. FR II sources are edge-brightened, probably because of shock heating as the radio-emitting plasma interacts with the ambient intergalactic medium. FR I sources are not edge-brightened, which suggests that their outflows are subsonic. Data courtesy of A. H. Bridle, figure by R. W. Pogge.

In addition to the compact and extended components, radio sources also often have features known as ``jets'', ¹ which are extended linear structures (Bridle and Perley 1984). An example of a jet is seen in the FR II source shown in Fig. 1.4. Jets appear to originate at the central compact source and lead out to the extended lobes. They often show bends or wiggles between the central source and the point where the jet appears to expand into the extended radio structure. The appearance of jets suggests that they transport energy and particles from the compact source to the extended regions. Jets often appear on only one side of the radio source, and in cases where jets are seen on both sides one side (the ``counter-jet'') is much fainter than the other. The difference in brightness is thought to be primarily attributable to `Doppler beaming' which preferentially enhances the surface brightness on the side that is approaching the observer.

The relative strength of the extended, compact, and jet components varies with frequency since the different components have different spectral shapes. The relative strengths also show considerable variation from source to source, with ``lobe-dominated'' sources having steep spectra and ``core-dominated'' sources having flat spectra. At least part of the observed differences among quasars must be due to orientation effects; whereas the extended components probably emit compact and jet components emit anisotropically. This will be discussed further in Chapters 4 and 7.

Because of the different observing frequencies and detection limits of the various radio surveys used to find quasars, each of the major radio-source catalogs contains certain selection biases. For example, the 3C sources tend to be those which are the very brightest at low frequencies, which will bias selection toward the high-luminosity, steep-spectrum (and thus lobe-dominated) sources. The 4C survey was more sensitive and tended to turn up sources which are in general of lower luminosity. On the other hand, the Parkes survey, which was undertaken at higher frequencies, tended to select a relatively greater fraction of flat-spectrum (and thus core-dominated) sources. Steep-spectrum sources, on the other hand, can be missed because they can fall below the detection limit at higher frequencies. For example, a source near the 4C detection limit (2 Jy at 178 MHz) will fall below the Parkes detection threshold (1 Jy at 1410 MHz) for alpha gtapprox 0.34.

¹ The term ``jet'' was first used in extragalactic astronomy by Baade and Minkowski (1954) to describe the optical linear feature in M87 that was first described by Curtis (1913). Despite the implication of the name, there is no unambiguous evidence that jets involve high outflow velocities, except in the innermost regions (§4.4.2). Back.