Active Galaxies

Types of Active Galaxies

Active galaxies differ from normal galaxies in a number of ways. These are outlined in the pages linked here.

All large galaxies seem to have super-massive black holes in their cores, but in active galaxies an accretion disk feeds material into the black hole while streaming matter and radiation outward. There are many types of active galaxies, and they appear differently to observers who are viewing them at different orientations and/or at different energies. The links below describe some of the properties of active galaxies.

BL Lacertae Objects (Blazars)

Bright, star-like objects generally lacking any emission lines.

BL Lac objects are named after the prototype object which was first believed to be a variable star in our galaxy. However, because of its similarities to AGN, BL Lac is now believed to be an extra-galactic object. Because of their occasional wild variability, these objects are sometimes referred to as “blazars.” This name also alludes to the similarities these objects have with quasars.

BL Lac objects have most of the following characteristics:

• Great and rapid variability in the radio, IR, and visible regions Optical variability of up to 4 magnitudes is not uncommon. This amounts to a change in brightness of nearly a factor of 20, and a few BL Lac objects have been reported to undergo changes in brightness of a factor of 100. While variations of 10 to 30% have been noted from night-to-night for some objects, the larger variations usually take place over months or years.
• No emission or absorption lines… continuous spectrum only
• Compact radio source with non-thermal continuous spectrum extending into the IR and visible
• Stellar appearing optical source with virtually no structure Some BL Lac objects reveal “fuzz” when observed with the largest telescopes. This “fuzz,” or faint nebulosity surrounds a point-like stellar appearing source.
• Strong and rapidly varying polarization

Image from: AAVSO – ask for permission (“Image of BL Lac, the prototypical “BL Lac” object. Credit AAVSO.”)

Approximately 40 BL Lac objects have been identified. Because of the virtual absence of emission or absorption lines, red-shifts are generally unknown. They are believed to be extra-galactic because of their radio properties and because of the “fuzz” observed surrounding some objects. The “fuzz,” when observed, seems to have a spectrum similar to that of an elliptical galaxy. A number of BL Lac objects are known in the vicinity of clusters of galaxies, so this provides indirect evidence in support of their extra-galactic nature.

Low Ionization Nuclear Emission Regions

Most spiral galaxies exhibit some weak emission lines due to the HII regions in the spiral arms. However, if emission lines are observed originating from the nucleus and not associated with the disk, then this is an indication of some sort of activity.

Objects with very low activity levels are sometimes termed LINERs (Low Ionization Nuclear Emission Region). The ionization energies are relatively small as indicated by the fact that lines like [O II] tend to be stronger than [O III].

(Image: LINER Galaxy NGC 4258, Image: Soren Larsen, UC Lick Observatories)

NGC 4258 Central Black Hole

The object NGC 4258 (also M106) is a particularly interesting example of a LINER because it was the first object for which nearly unassailable evidence for a black hole was discovered. This was due to a fortunate set of circumstances in this system. First, it had the proper physical conditions in its nucleus to produce MASER emission (like a LASER, but in the microwave rather than the visible part of the spectrum). Second, the Earth happens to lie in the plane of the disk surrounding the black hole, so we see the MASER emission.

The optical image above shows a fairly normal looking spiral. However, on closer inspection you might notice that the spiral arms look slightly disturbed. The galaxy also has a very bright nucleus, which causes it to be classed as a Seyfert galaxy.

At left are radio and HST images. The MASER sources are seen in the radio emission. These sources are plotted at their position in the disk below the two images.

Image courtesy of NRAO/AUI and L Greenhill (Harvard-Smithsonian Center for Astrophysics)

Left is an artist’s interpretation of what this system might look like. Note the warped disk, seen in the MASER emission above. Note also the two jets emanating from the nucleus. We do not see these jets at all, but infer their existence because we think they are typical of AGN behavior. The MASER emission is plotted below the point in the disk where it is observed.

Image courtesy of NRAO/AUI and Artist: John Kagaya (Hoshi No Techou)

For additional information on this system and how the presence of the black hole is inferred, go to the NRAO page on the object.


In the late 1950s radio astronomers began compiling catalogs of radio sources. At that time, there were quite a number of radio sources which were not associated with previously known optical objects. The primary difficulty was that radio telescopes had poor resolution and large beam sizes so that it was often not possible to identify which of as many as several hundred faint stars was actually associated with a strong radio source. (Note that today this would not be a problem due to the advancements in resolution brought about by radio interferometric techniques). In 1960 one unassociated radio source (3C 48) was finally identified with a 16th magnitude stellar object. The identification was made when the moon occulted the radio source. Since this relatively strong radio source was associated with an object that appeared stellar, this category of objects became known as quasi-stellar radio sources (QSRS). This is the origin of the name quasar.

HST image of 3C 273. Note the starlike appearance. Note also the jet of material being ejected toward the lower right.

When a spectrum of this faint source was obtained it was observed to have strong broad emission lines which could not be identified with emission lines of known chemical elements. In 1963 the strong radio source 3C 273 was also identified with an inconspicuous 13th magnitude “star.” This object too exhibited strong emission lines which could not be identified. Eventually a sequence of emission lines similar to the Balmer series of hydrogen was recognized in the spectrum of 3C 273, but shifted far to the red. The red shift observed for 3C 273, z = 0.158, was greater than the highest known redshift for a galaxy. However, assuming this redshift, the remaining emission lines could readily be identified with the standard emission lines known. Because the emission lines identified in this way are the same lines known to be present in nearby galaxies, and because galaxies are the only objects known to have systematic large redshifts, quasars are generally believed to be extragalactic.

Discovery spectrum of 3C 273. Note the redshift of the hydrogen lines in the source relative to the comparison spectrum. The wavelengths of H-Delta, H-Gamma and H-Beta are, respectively, 410, 434 and 486 nm.

When Maarten Schmidt measured the spectrum of 3C 273 in 1964 it was the most distant object known… for about 15 minutes! Then his colleagues in the next office measured the spectrum of 3C 48, which gave z = 0.36.

If quasars are assumed to follow the Hubble law for distances, the large redshifts must correspond with large distances. Indeed, at the large distances implied by some quasar redshifts, a normal galaxy would be too faint to be observed. Thus, quasars would appear to be hundreds of times brighter than normal galaxies.

Quasars have now been observed with redshifts ranging from 0.06 to above 6.0. However, there are no nearby quasars, and the number of quasars increases with redshift. Since redshift is related to distance, objects with large redshifts are farther from us; it has taken their light longer to reach us, and so we see them from a time farther into the past. Since quasars are more numerous at high redshift, they must also have been more common when the universe was younger. Quasars have either disappeared during the present time or have changed their characteristics and evolved into less luminous objects. According to current thinking on the subject, all large galaxies went through an active “quasar” stage early in their existence, but have now settled down into quiet middle age. In support of this idea is the growing evidence that all large galaxies contain supermassive black holes at their cores… the same type of object thought to power quasars and other AGN.

Radio Galaxies

Nearby galaxies (including the Milky Way) do radiate in the radio region, but they are generally weak sources. Typical radio emission from normal galaxies falls in the range 1030 to 1033 watts, with spirals falling at the bottom end and ellipticals at the upper end (for comparison, the Sun emits around 1026 W if all wavelengths are included). Some galaxies, however, are far more luminous in the radio region. Low energy AGN such as Seyferts, starbursts and active elliptical galaxies can be up to 1000 times more powerful in the radio than the most powerful normal galaxies. At the most luminous, quasars and radio galaxies can emit upward of 1038 W in the radio.

(Image: Cygnus A radio galaxy. Image R.A. Perley, J.J. Cowan, J.W. Dreher, and NRAO/AUI/VLA)

Radio galaxies appear to fall naturally into two types: compact and extended. Compact sources show no structure, while extended sources often show a double lobed structure. It is, of course, possible that at least some of the compact sources are simply too distant to reveal any structure.

In general, extended radio galaxies have radio emission from a region that is far larger than the optical image of the galaxy. These sources often show a double structure consisting of two gigantic radio lobes on opposite sides of the nucleus of the galaxy. The lobes can be separated by many kiloparsecs, or even megaparsecs. The optical galaxy generally lies near the middle of a straight line connecting the radio lobes. Typically the nucleus of the optical galaxy is itself a weak radio source, and the lobes dominate the radio emission. Occasionally jets are observed in the optical, in X-rays, and in the radio. These jets are pointing directly toward the radio lobes. Such sources are termed “classical double” sources.

In terms of structure, extended radio galaxies may be characterized into three main groups.

Classical Doubles

These have the highest radio luminosities and have the lobes aligned through the center of the galaxy. There are bright hot spots at the ends of the lobes. (like Cyg A, see images at right and below)
Image: Cygnus A Radio galaxy, shown in optical with radio contours overlain.


Bent Tails

These have intermediate luminosities, a bend through the nucleus, and tail-like protrusions. (like Cen A)

Narrow Tails

These have the lowest luminosities, U shapes, and are (usually) rapidly moving through a galaxy cluster.

(Left) Centaurus A radio galaxy. Image: NRAO/AUI/VLA and NASA/ESA Hubble Space Telescope
(Right) Radio Galaxy NGC 1265. Image courtesy of NRAO/AUI and C. O’Dea & F. Owen.

To learn more about radio galaxies, read Chapter 13 in Galactic and Extragalactic Radio Astronomy, Ed. by Verschner, G. L., and Kellermann, K. I., Springer-Verlag, 1988. Sadly, the book is now out of print, but you can probably find it in a university library. The short synopsis here was in large part taken from there.

Seyfert Galaxies

Seyfert galaxies were first identified by Carl Seyfert in 1943. Initially they were classified as galaxies with broad emission lines. These galaxies also have unusually bright nuclei. Seyfert initially identified 6 such objects, and there are now approximately 90 known. They all appear to be spiral galaxies, and although 5-10% might be elliptical, the questionable objects all have extremely small angular sizes so it is difficult to discern any morphology. It appears that about 1% of all spiral galaxies might be Seyferts. It has been hypothesized that at least some spirals (or possibly all spirals) go through an active Seyfert phase under the proper conditions. Since most Seyfert galaxies appear to be in close binary galaxy systems, it may be that tidal interactions induce or even turn on the Seyfert phenomenon.

Since Seyferts generally exhibit both permitted and forbidden lines, the emission must be produced in different regions of the galaxy where different physical conditions prevail. Forbidden lines can only be produced in low density regions, whereas higher densities must occur where permitted lines are produced. In terms of their spectra there are now three sub-types of Seyfert recognized.

Three subtypes of Seyfert galaxies
Type 1
These are the “classic” Seyfert galaxies. They have extremely broad permitted lines of hydrogen. If the broadening is interpreted as being due to Doppler motions, the broadening corresponds to velocities of up to 5,000 or 10,000 km/sec! The forbidden lines (typically O II, O III, N II, S II, etc.) are only “moderately” broadened by 200-400 km/sec. For comparison, in an ordinary galaxy the escape velocity from the galactic disk is normally less than 300 km/sec. Of course, the emission lines in a Seyfert originate in the nucleus of the galaxy.
Type 2
These Seyferts have “narrow” lines by comparison to the Type 1’s. The broadening rarely amounts to more than 200-400 km/sec (1000 km/sec in one case) and the permitted and forbidden lines have the same widths.
Type 1.5
Some Type 1 objects are observed which have both broad and narrow components for the permitted hydrogen lines. A narrow emission peak seems to extend above the broad underlying emission line.


Example spectra of Type I and Type II Seyfert galaxies. Upper spectrum is actually more like a Type 1.5 (Note the narrow lines extending above the core of the broad lines). Image from Bill Keel’s Slideset on AGN.

The standard model for a Seyfert galaxy involves three components. First, a tiny central source of high energy ionizing photons, and then two distinct surrounding regions with different gas densities. Presumably, the inner region is the Broad Line Region (BLR) with high densities appropriate for the production of permitted lines. The velocities in this region must approach the 5,000-10,000 km/sec values deduced from the line widths. Because the broad lines are observed to undergo significant variability over periods of weeks or months, the size of this region cannot be much greater than a light month or so. This size corresponds to about 1011 km and is not much larger than the size of our planetary system. Gas densities in the BLR must be on the order of 1013-1015 ions/m3. Outside the BLR must be the Narrow Line Region (NLR) where the gas densities are low enough to allow forbidden line production. The scale size of the NLR must be about 102-103 times larger than the BLR. There are no observational reports of variability in the narrow lines.

Since there is no reason why permitted lines cannot be produced in the low density NLR, the existence of Type 1.5 Seyferts is totally understandable. Furthermore, slight differences in the widths of the narrow lines suggests that the density in the NLR decreases with radial distance from the nucleus. The critical densities for forbidden lines from different atomic species vary, and those lines coming from the least dense regions exhibit somewhat narrower lines.

The continuous spectra of Seyferts seems to be a combination of stellar, nonthermal, and IR emission from dust. Seyfert galaxies are not strong radio sources and the most sensitive radio surveys have detected only about half of the known Seyferts. Likewise, Seyfert galaxies are not strong X-ray sources. (????)

Starburst Galaxies

Some galaxies appear to have experienced intense episodes of star formation between 10 and 100 million years ago. In some cases we see the galaxies while the star formation is still occurring. Strong emission lines, generally associated with HII regions, are often detected in the nuclear regions and throughout the disk. These galaxies are intense IR sources and relatively weak radio and X-ray sources. It appears that a large concentration of massive type-I objects (like OB stars, massive X-ray binaries, and supernovae) are heating the extensive dust concentrations and producing the intense IR emission. The rate that these galaxies convert their gas and dust to stars can be 10 to 100 times higher than a typical star forming galaxy like our own.

Starburst galaxies are usually very dusty late-type galaxies. The increased star formation rate has been triggered by tidal interactions with a close massive companion galaxy. This can be seen in the nearby M81/M82 system, shown in the HI (21 cm) image, in the middle of the pictures at left. In the optical image, there is no indication of an interaction between the members of the group, but in the radio it is obvious that these galaxies have recently influenced one another.

The M81 system shown in optical (left, from theDigitized Sky Survey) and HI 21 cm radio. Note the extent of the HI throughout the system. M81 is the large spiral, M82 is above it. The small galaxy to lower left is NGC 3077. Image courtesy of NRAO/AUI.

Image courtesy of NRAO/AUI.

In other starburst systems the collision is still well underway, probably leading to an eventual merging of the systems involved. This type of system is shown in the image of the Antennae galaxies, at right bottom. It is possible that this merger/starburst scenario is the fate that awaits the Milky Way and its companion M31 in several billion years. Afterward, instead of two grand spirals, the Milky Way and Andromeda will have merged to form a single large elliptical galaxy.

The Antennae galaxy merger system is typical of those that form starbursts. Image: NASA/ESA Hubble Space Telescope.

Another common property of starburst galaxies is a strong galactic wind, presumably driven by the cumulative effect of many supernovae occurring in rapid succession. Whereas a normal galaxy like the Milky Way will have one or two supernovae per century, in a starburst the supernova rate can be a hundred times as much. The image at left shows gas being ejected from M82. The purple material is hydrogen leaving the galaxy at several hundred km/s.

This image of M82 shows a massive gas outflow (the purple material along the galaxy’s axis). The image is taken in the light of the Hydrogen-Alpha emission line, and so shows hot, ionized hydrogen gas. Massive outflows. Image: NASA/ESA Hubble Space Telescope.

Occasionally galaxies are found which show the signs of recent star burst activity. Their spectra indicate a normal elliptical type galaxy with an old population of stars and no gas, dust or star formation. However, these galaxies contain A stars, and so are referred to as E+A galaxies. Elliptical galaxies should not contain A stars, which are too short lived to be present in a normal elliptical. A clue to the nature of these galaxies is that they are often surrounded by extend halos of diffuse gas (which can sometimes be seen in absorption against a quasar or other background object); the gas is the remnant of the galaxy’s interstellar medium, now blasted into intergalactic space. It could be that M82 is headed for this fate in another billion years or so.

AGN Characteristics

Galaxies with active nuclei (AGN), classically exhibit all or most of the following characteristics:

  1. High luminosity, greater than 1037 W
  2. Nonthermal radio source (synchrotron) with excessive UV, IR, radio, and X- ray flux as compared to an ordinary galaxy
  3. Small region of rapid variability, exhibiting significant variations over time periods of months or less, in the galactic nucleus
  4. High brightness in nuclear region as compared to other regions of the galaxy
  5. Jet-like protuberances
  6. Broad emission lines (sometimes)


Processes in Active Galaxies – Synchrotron Radiation

The emission from ordinary galaxies is dominated by thermal radiation from stars and dust. Active galaxies are dominated by non-thermal processes like synchrotron radiation and thermal processes characterized by exceedingly high temperatures not found in stars (i.e. large energies). Below we outline some of the character of each.

Thermal Emission

In normal galaxies, the temperatures found usually vary from about 10 K for the coldest dust and gas in the interstellar medium to several thousand kelvin for stellar photospheres. Stars generally fall in the range between 3000 and 30000 K (the sun has a surface temperature near 6000 K), and so they emit most of their energy near the visual wavelength region. The type of radiation emitted by stars is called Planck radiation, or sometimes black body radiation. Max Planck first explained such radiation from a theoretical basis in 1900. He found that the distribution of energy with wavelength of such an object depends on temperature alone, and not on composition, size, phase or other properties. This is why Planck emission is referred to as a type of thermal emission (there are others as well). Of course, black body emitters are an idealization, and stars are really just very good approximations to them. Nonetheless, the Planck radiation law predicts extremely well that the hottest stars will have their peak emission in the UV while the emission of the coolest stars will peak in the near to mid-IR. Example spectra for Planck emitters are given below (at right) for three temperatures: 3000K, 6000K and 30,000K.

The Wien displacement law describes the relationship between the temperature of a Planck emitter and the wavelength of the peak of the Planck curve describing its spectrum. The table below lists the peak wavelength of the Planck emission for four temperatures, corresponding to the Planck curves in the figure to the right. Note that the curve for 100K falls outside the boundaries of the plot and so cannot be seen.

max (in microns) = 2898 / T

Here is the temperature in Kelvin

T (K) lambda max (microns)
100 28.98
3,000 0.9660
6,000 0.4830
30,000 0.09660

At wavelengths longer than the peak the radiation emitted decreases slowly with increasing wavelength. (At wavelengths shorter than the peak there is a rapid decrease.) At radio wavelengths (or frequencies) we are observing on the long wavelength tail of the energy distribution for ordinary stellar and galactic sources. Thus, in the radio region, ordinary astronomical sources which are thermal emitters are brighter at higher frequencies (shorter wavelengths).

Synchrotron Emission

Synchrotron radiation is produced when energetic charged particles (electrons in this context) move through a region of space containing a magnetic field. The motions of the particles are deflected by the magnetic field, causing them to gyrate around the magnetic field in a plane perpendicular to the direction of the field lines. Since accelerating charged particles causes them to radiate (according to the Larmor radiation formula), these particles will radiate as they gyrate around the magnetic field. When the energy distribution of the radiating particles follows a “power law” (see below) which is often true, the spectral distribution of this radiation at long wavelengths is described by the following relation:




Here s is called the spectral index of the radiation, the Greek letter nu is the frequency and F is the flux. This type of relation is called a power law relation because the flux varies with a constant power, s, of the frequency. The equation relates the flux at some particular frequency nu (ν), to that at a reference frequency ν0. The spectral index is related to the index describing the energy distribution of the electrons producing the radiation, though it is not the same. Radio galaxies and quasars have similar radio properties and have spectral indices between s = 0.7 and s = 1.2. Compact radio sources have a flatter spectrum and tend to have a spectral index near s = 0.4.

The image at right shows the synchrotron spectrum for the Milky Way. At short wavelengths (high frequencies) the emission has a power law index s = 0.6. Note the turnover in the power law at low frequencies. This is caused by a turnover in the number of relativistic electrons at low energies. Figure from Cummings, Stone and Vogt, 13th International Cosmic Ray Conference, Denver, 1973.

Processes in Active Galaxies – Emission Lines

Emission lines are produced when an excited atom de-excites, i.e., when an electron in such an atom transitions from a high-energy orbital to an available lower energy orbital. Collisions in a high temperature gas are generally sufficient to provide the excitation required, provided that further collisions do not de-excite the atoms before they radiate. Thus, for emission lines to be present the gas must typically have a low density in addition to a high temperature. This condition is not a particularly stringent one, and when gases are heated sufficiently they generally will show emission lines, with the exception of the “forbidden” lines discussed below.

If temperatures are high enough the atoms can actually be ionized: an electron or electrons can be stripped completely off the atoms. Atomic ionization can also be the result of absorption of high energy photons (as in H II regions). In either case the emission lines result when an ion recaptures a free electron, which then cascades down through the ionic energy levels. The spectrum from the ionic form of an atom is always different from the spectrum of the neutral atom, and different ionization states also differ from each other in their spectra.

As has already been mentioned, different sets of emission lines are possible depending upon the density of the gas. In low density regions, where collisions are less frequent, some lines are seen which are never seen in high density regions. For historical reasons these have been termed “forbidden” lines. The lines are not really “forbidden”; they originate in transitions between so-called “meta-stable” atomic states and lower energy states. These meta-stable states are much longer-lived than other atomic states. In a high-density gas, atoms tend to be collisionally de-excited from meta-stable states before they can radiate. For that reason forbidden lines are not observed under normal laboratory conditions on the Earth. Square brackets [ … ] are normally used to indicate forbidden lines.

The table below has a partial list of emission lines often seen in active galaxies. Many of these same lines are also seen in star forming regions in galaxies, the exception being the lines of high ionization (N V and C IV are examples). At right are examples of emission line spectra from two Seyfert galaxies. Spectra from Bill Keel’s Slideset on AGN.

H I (Lyman Alpha) 121.6 N V 124.0 C IV 154.9 [C III] 190.9
Mg II 279.8 [O II] 372.7 [Ne III] 386.8 H-Delta 410.2
H-Gamma 434.1 H-Beta 486.1 [O III] 495.9 [O III] 500.7
[S II] 671.7 [N II] 654.8 H-Alpha 656.3 [N II] 658.4
[S II] 673.1