§2 Active galactic nuclei (AGN)

By the 1960's, astronomers began to realize that there were unusual signs of activity in the centres of many galaxies. These galaxies had intense concentrations of blue light, quite uncharacteristic of the standard radiation received from aggregates of stars and gas, which are the normal sources of radiation in a galaxy. The spectrum of this new source of light contained too much blue and ultraviolet radiation to come from even the hottest and most massive stars. In some cases, these sources of radiation were found to be as bright as the host galaxy, their emission often varying with time. Galaxies containing these central sources came to be known as Active Galactic Nuclei (AGN). Observations made at radio frequencies revealed that certain classes of these galaxies contained narrow, fast jets of gas as was briefly mentioned in the previous section.

It is likely that all galaxies posses some form of activity in their nuclei associated with a supermassive black hole. The luminosities of the nuclei can range from $\unit{ \power{ 10 }{ 40 } }{ erg \usk \reciprocal \second }$ to $\unit{ \power{ 10 }{ 47 } }{ erg \usk \reciprocal \second }$.

The classification of AGN is a very confused and confusing subject (Robson, 1996). This is mostly because of observational problems. It is not possible to obtain full spectral coverage for all objects and so, it is not easy to reconcile a classification based on say, X-ray properties with that of optical emission lines of a different sample. Other problems are historical in origin. The phrase ``quasi-stellar'' object is an example of this. One of the challenges of AGN research is to infer the physical processes that take place in different AGN assuming that they are simple, in spite of the confusing character of the observations (Blandford, 1990).

Let us describe briefly one of the ``modern'' classification schemes. The observational details of these observations can be found in Woltjer (1990) and Robson (1996), and are clearly summarised by Begelman et al. (1984), Blandford (1990), Longair (1995) and Reynolds (1996).

AGN can be divided into two categories, radio-loud objects, with a luminosity $L_{\unit{ 5 }{ \giga \hertz } } \gtrsim \unit{ \power{ 10 }{ 24 } }{ \watt \reciprocal \hertz \reciprocal \steradian }$, or radio-quiet, for which $L_{\unit{ 5 }{ \giga \hertz } } \lesssim \unit{ \power{ 10 }{ 24 } }{ \watt \reciprocal \hertz \reciprocal \steradian }$, objects on the basis of their radio properties. About $\unit{ 10 }{ \% }$ of all AGN are radio-loud. The rest are radio-quiet.

Radio-loud AGN emit collimated beams or jets of plasma which feed energy and particles to the three dimensional dumbbell structures or lobes (Blandford, 1990; Begelman et al., 1984). The jets and the lobes are sources of continuum radio emission due to synchrotron radiation from relativistic electrons spiralling in a magnetic field in the jets and lobes.

The flow inside the jets in radio-loud AGN may expand, or propagate, with relativistic velocities. This is inferred from the following two observations. Firstly, high spatial resolution radio mapping, with Very Long Baseline Interferometry (VLBI), made at epochs separated by years show that there are blobs within the jet moving outwards from the nucleus. In many cases the proper motion of these blobs in the sky show that the speeds exceed the speed of light. These so-called superluminal motions were first predicted by Rees (1966) and they are produced as a result of relativistic motion of the blobs moving at small angles to the line of sight to the observer (Blandford, 1990). Superluminal motion has been the most successful interpretation of these observations, however many other unsuccessful models have been developed to account for it (see Recami et al., 1986, for a review of different models). Secondly, many radio-loud AGN objects, particularly FR--II radio galaxies and radio-loud quasars as defined later, display only one jet. It is very difficult to believe that only one-sided jets are created in symmetrical double radio sources. In fact, it is believed that there are two jets, but due to the relativistic bulk motion of the flow within the jet, there are significant beaming effects due to aberration (Blandford, 1990; Begelman et al., 1984). This is believed to be the reason why counter-jets are not observed on the radio maps. These observations imply that the bulk motion of the plasma inside the jet has a Lorentz factor $\gamma \! \sim \! 3$-$10$.

Radio-quiet AGN do not show large scale kilo parsec jets. However, some small parsec jets have been observed in some of these sources. An outflow of bipolar radio emission, extending less than $\unit{ 5 }{ \kilo pc }$, is very often seen from these AGN. There are some larger scale structures associated with this objects, but their nature is rather obscure and they might be the result of an AGN outflow or a starburst driven superwind coming from the host galaxy.

Radio-quiet AGN can be classified on their basis of optical/UV spectral properties. Briefly, this classification is as follows:

Seyfert I galaxies / radio-quiet quasars.- Their properties are complex and multi-component. They are often referred as broad-line or type-1 AGN because they have broad permitted optical/UV emission lines (with full width at half maximum $\textrm{FWHM} \! \sim \unit{ 200\text{-}20000 }{ \kilo \meter \usk \reciprocal \second }$) in the nuclear spectrum. The region from which all this emission arises, the so called broad line region (BLR), contains dense photoionised clouds (with an electron number density $n_\mathrm{e} \! \sim \unit{ \power{ 10 }{ 9 } }{ \centi \meter \rpcubed }$) with a small volume filling factor. The optical/UV spectra of these objects shows also narrow permitted and forbidden emission lines ( $\textrm{FWHM} \! \sim \! \unit{ 500 }{ \kilo \meter \usk \reciprocal \second }$). The gas in this region, the narrow line region (NLR), is more tenuous that in the BLR and lies at greater distances from the core of the AGN. Traditionally, the distinction between a Seyfert I galaxy and a radio-quiet quasar, often referred as quasi-stellar object or QSO, was made on the basis of whether the galaxy or AGN was discovered first.

Seyfert II galaxies.- The spectra of these active nuclei show narrow permitted and forbidden emission lines very similar to those of Seyfert I galaxies. However, Seyfert II galaxies show no broad lines and the optical/UV continuum radiation is much weaker.

Radio-loud AGN can be sub-divided on the basis of their morphology and optical/UV spectral properties. The strong extragalactic radio sources have been divided mainly into two categories, the extended radio sources, the bulk of their emission originating from regions more than a kiloparsec from the nucleus of the associated galaxy or quasar, and the compact radio sources, in which the emission comes from a region less than a kiloparsec from the nucleus. The extended sources have a steep spectrum $\alpha \! \sim \! 0.5$-$1$ , whereas the compact sources have a flat one $\alpha \! \sim 0$-$0.5$. The majority of the extended radio sources are associated with elliptical galaxies and the minority with quasars. The extended radio sources can be divided into two main categories:

Fanaroff-Riley type I radio galaxies (FR I).- These sources are also called edge-darkened sources because the radio surface brightness profiles fall continuously as the distance from the nucleus towards the edge of the radio lobe increases. They are low luminosity objects with a luminosity $L_{\text{radio}} \! \lesssim \unit{ \power{ 10 }{ 42 } }{ erg \usk \reciprocal \second }$, and usually have two anti-parallel jets emerging from the nucleus of the source. FR I sources show narrow emission lines coming from a NLR, but they do not show broad optical/UV lines.

Fanaroff-Riley type II radio galaxies (FR II).- The radio surface brightness of these sources increase as one moves towards the outer edge of the halo, where the jets terminate in strong shocks. As a result of this, FR II radio galaxies are often referred as edge-brightened sources. These radio sources are luminous ( $L_{\text{radio}} \! \gtrsim \! \unit{ \power{ 10 }{ 42 } }{ erg \usk \reciprocal \second }$) and usually have at most one jet with spectral index $\alpha \! \sim \! 0.5$. They show linear polarisation with the electric vector perpendicular to the jet. The optical characteristics of this type of AGN are very similar to Seyfert I galaxies, showing broad and narrow emission lines with a strong continuum, or like Seyfert II galaxies, showing only narrow lines. When Seyfert I-like, these object are classified as broad-line radio galaxies (BLRG) and when Seyfert II-like as narrow-line radio galaxies (NLRG).

Compact radio sources are mainly divided in two categories:

Radio-loud quasars.- The differences between BLRG and radio-loud quasars (RLQ), also known as quasars, is blurred. The distinction is largely made on the basis of whether or not the active nucleus overwhelms the light from the galaxy. If this happens the object is called a RLQ. These RLQ objects often have one-sided jets in which superluminal motion is observed. These sources, sometimes called core-jet sources have a flat spectrum, $\alpha \! \sim \! 0$, core and a somewhat steeper spectrum in the one-sided jet.

Blazars.- These are radio-loud objects exhibiting very strong and variable continuum emission at all wavelengths and they have a high optical linear polarisation. The emission lines in these sources are either very weak or absent in the optical/UV spectrum. When these emission lines are present, the objects are classified as optically violent variable (OVV) quasars and when these lines are absent the sources are referred as BL Lac objects, after the first example of this last to be discovered.

Footnotes

...

The spectral index $\alpha$ is defined by $\mathsf{S}_\nu \propto \nu^{-\alpha}$, where $\mathsf{S}_\nu$ is the flux density at a frequency $\nu$.