Table of Contents

• Introduction
• Frequently Asked Research Questions
  
• Classification System
• Central Black Hole
• The Accretion Disk
• The AGN Continuum
• Broad Emission Lines
• Narrow Emission Lines
• Broad Absorption Lines
• Radio Jets
• AGN Hosts/Environment & Fueling
• AGN Evolution

Introduction

What are Active Galactic Nuclei?

As astrophysics approached the close of the 20th century, we learned that many of the galaxies that we can observe in detail seem to have very massive black holes in their cores. For instance, our galaxy has a black hole of mass approximately 2 x 10^6 solar masses at its core, and studies of gas and stellar motion in other galaxies have revealed other supermassive (with masses greater than 10^6 solar masses) black holes in their cores. Most of these galaxies don't exhibit any clue to their central black holes besides the stellar or gas motion used to find their masses. However, some galaxies (for reasons we don't quite yet understand) have black holes that draw much more attention to themselves by generating extraordinary amounts of energy in their host-galaxy cores: much more than ordinary galaxies. We call these galaxies Active Galactic Nuclei, or AGNs, for short. So, at the most basic level, AGNs are simply the cores of galaxies that are not quiescent like normal galaxy cores: for instance, some of them emit an amount of energy comparable to the output of an entire galaxy from their cores, which are regions measuring only parsecs in size.

This sounds pretty simple, but the study of Active Galactic Nuclei (AGNs) sometimes becomes very confusing because there are a wide range of ways that the cores of galaxies (galactic nuclei) can appear active, or at least "not normal." As Krolik (1999) points out, it's almost as if there's a "menu" of extraordinary activity options that AGNs can select from. For instance, some emit significantly more radio energy than others, some have much broader emission lines in their spectra, and on and on. Each time a new combination of these properties are found, we "see" a new type of AGN. Therefore, it's not hard to see why we now have a wide range of different AGNs as we try to group them by their different activities.

A beautiful, organized chart of all of the different kinds of AGNs, their relative populations, and how they are related is shown in Figure 1.1 of the Introduction to Dr. Roy's Thesis

Where do they come from?

AGNs become active because gas is being accreted onto the central supermassive black hole. Thus, it seems reasonable to suggest that any process that can bring gas from the galaxy (from a kiloparsec away, in the large-scale structure of the galaxy) down to the central black hole may cause a galaxy nucleus to become active. This may not be as easy as it sounds, as some researchers now think there may be different processes that allow the transport of gas to the central regions of the AGN. However, galaxy-galaxy collisions may play a key role in disturbing a galaxy's gas distribution so much that gas falls into the center of the galaxy, and AGN activity is triggered.

Where do they go?

After the gas in the central region of the AGN has been accreted into the central black hole, we believe that the AGN "shuts down." Therefore, after such activity, the AGN stops outputting energy and may become, for all intents and purposes, a normal galaxy.

• "An Introduction to Active Galactic Nuclei" by Bradley M. Peterson
• "Active Galactic Nuclei: From the Central Black Hole to the Galactic Environment" by Julian H. Krolik
• "Quasars and Active Galactic Nuclei : An Introduction" by Ajit Kembhavi & Jayant V. Narlikar
• "Astrophysics of Gaseous Nebulae and Active Galactic Nuclei" by Donald E. Osterbrock
• "Active Galactic Nuclei: Saas Fee Advanced Course 20 (Lecture Notes 1990 Swiss Society for Astrophysics and Astronomy)" by R.D. Blandford, H. Netzer, L. Woltjer

• Pat Hall's Minicourse on AGN
• Bill Keel's WWW Gallery - Active Galaxies and Quasars

Frequently Asked Research Questions

• Does it work?  Most papers say that yes, the Unified Scheme fits their observations, but not all observations bear it out.  There are people on both sides:
• Yes: see astro-ph/0406111, for example.
• The Unification Scenario does not seem to work in the following cases:
• some quasars seem to be obscured optically, but have very little obscuring column at X-ray wavelengths.
  
• astro-ph/0406163 shows possible evidence for AGN that don't have any broad emission lines and don't appear to be obscured. 
  
• There is also some evidence that Sy1 and Sy2 galaxies have different radio brightnesses (Roy et al., 1994, ApJ 432:496), but this has recently been contested (astro-ph/0406611) in a study where no significant difference in radio emission was found.
• astro-ph/0408469 reports that isophotal twists are twice as common in Sy2 as in Sy1 galaxies.  They propose that perhaps this is evidence for evolution between the types.
  

• What is the quasar population as a function of time?  It seems to be different for quasars of different luminosity (see astro-ph/0406330).  High luminosity quasars (log L_x > 44.5) appear to have peaked around a z of 2 (+/- 0.5 or so), while lower luminosity AGN seem to peak a progressively lower z.  This may perhaps be due to less vigourous fueling of the AGN as the galaxy merging rate slows?
  
• Is the same physics at work in AGNs and microquasars? A good review of this idea is in astro-ph/0405433.
• Does the occurance of AGNs depend on environment?  Certainly the occurance of AGNs depends on pulling gas from the outer, kpc-scale regions of the disk to the inner regions, but it's not at all clear how this is done (see astro-ph/0407068 and astro-ph/0408282 for very nice overviews).  The occurance of AGNs also does not seem to depend on mergers or galaxy-galaxy interactions (de Robertis, Yee, & Hayhoe 1998), but this is still debated (Krongold, Dultzin-Hacyan, & Marziani 2002).  [recent work: Best (2004) (radio-loud AGN seem to prefer clusters and the ratio of emission-line to absorption-line AGN changes as a function of environment)]
• How does gas reach the central BH?  This is a key question, and appears to happen in stages, as if there are different physical processes that open "gates" at different, characteristic radii and on different timescales. It appears that Seyfert galaxies may occur slightly more often (at 2.5 sigma) in barred galaxies rather than non-barred, but this isn't greatly statistically significant, and besides, sub-kpc nuclear bars would still be required to bring the gas into the core of the AGN, and those don't seem to be any more prevalent in Seyferts or non-Seyferts. 

• A good overview of quasar continuum emission across the entire spectrum is Risaliti & Elvis, astro-ph/0403618.
  
• Where does the soft X-ray excess come from? Papers commonly model this as blackbody emission with energies of a few hundred eV, although acknowledging that the accretion disk could not produce such high temperature emission.  It's been hypothesized that this emission could be the result of Comptonization of lower-energy photons by a higher energy `corona' above the accretion disk. It's also been suggested that the shape of the soft X-ray excess may be significantly affected by absorption (which may account for its variability?; astro-ph/0407045).
  
• Where does the hard X-ray (E > 2 keV) spectral component come from?  This is usually modeled as a hot corona above the accretion disk, which, via Inverse Compton scattering, upscatters optical/UV photons from the disk to X-ray energies.
• What forms the Fe K-alpha line?  Most models cite a small fraction of the hard X-ray continuum "back-illuminating" the disk, exciting the ions to form the Fe K-alpha line.  The same hard X-rays may also Compton down-scatter to produce a "Compton Reflection Hump" at high energies.  The location of the Fe K-alpha line is still intensely studied, since narrow Fe K-alpha seems to be the rule rather than the exception (see Bianchi et al. 2004, A&A 422, 65); no observations can yet resolve these narrow lines, so it's difficult to tell what their broadening is, and therefore, where they're produced, although most of the emission seems to come from neutral Fe in Compton-thick material, and in most Seyferts, the line is not broadened, so it must lie beyond approximately 20 R_g.
  
• Is there a Big Blue Bump in the UV?  If so, what produces it?
• What kind of dust do AGNs have?   It's suspected that AGN dust is different from the dust found in the Milky Way.  Some suggest (Hopkins et al. 2004, AJ, 128, 1112) that most AGNs have SMC-like dust, while others have proposed "gray" dust (at least in the UV; see Gaskell et al. 2004).
  

• How do single-peaked emission lines form if the gas is accreting in a disk?  A model using the disk wind's opacity to form single-peaked lines has been proposed (Murray & Chiang 1997) as well as a cloud model that depends on the optical depth of the individual clouds (Bottorff et al. 1997).  Recent work on this includes astro-ph/0405447, which proposes two separate regions for the BLR for producing the line cores (a spherical wind) and wings (an inner accretion disk).
  
• When double-peaked emission lines form, what is the source of their variability?
  
• What's the geometry of the Broad Line Region?  Is it made up of clouds or a continuous wind, or both?
• Can a continuous wind from the Broad Line Region act as both the source of the broad emission and absorption lines?

• X-ray variability: Sy1 galaxies display rapid X-ray variations, with no period, and we don't understand where this variability comes from.  As at other wavelengths, X-ray variability is studied by producing a Power Spectral Density curve, showing the degree of amplitude variability as a function of frequency.  For Sy 1 galaxies, generally, stronger variability is exhibited at larger timescales, and it appears that variability amplitudes are anti-correlated with X-ray luminosity and black hole mass (Markowitz & Edelson, astro-ph/0408045), although the amplitudes may saturate at the largest black hole masses and largest timescales.
  

• What's the source of the absorbing gas?  The Murray, Chiang, Grossman & Voit (1995) model predicts that the broad-absorption-line gas is simply the broad-emission-line gas, but further from the central source, and crossing the observer's line-of-sight in Broad Absorption Line QSOs.
  
• How is that gas accelerated?  There are various perspectives on this; models usually postulate radiative acceleration or magnetic (magneto-centrifugal) acceleration.
• Why do we see Narrow Line Seyfert 1 (NLS1) galaxies, with permitted lines only slightly broader than forbidden lines? Originally, there were two theories on this: one, that NLS1 might be Seyfert 1 objects but just viewed closer to the spin axis or perhaps have obscured BLRs, or two, that NLS1 galaxies have intrinsically lower mass black holes (10^6 to 10^7 solar masses), which must therefore accrete at near-Eddington or super-Eddington rates to maintain an AGN-like luminosity (Boller et al., 1996).   Mathur (2001) proposed, in line with this, that NLS1s are just young S1s.  The idea that the BLR is seen nearly face-on or obscured seems to have been discredited with the lack of any polarization signatures, so most researchers seem convinced that NLS1s are S1s with smaller black holes.  (Recent research on this: Botte et al. 2004 [showing NLS1s have lower M_BH and L_bulge, and follow M_BH-sigma])
  

• What are the radio jets made of?  This still seems to be a hotly contested question: some people favor an electron-positron composition, while others believe jets are made of electrons and protons. 
  
• Are the enhancements in emission in the jets made up of blobs of gas, or are they shocks? There seems to be no agreement on this.  A recent survey (astro-ph/0406116) of radio jets seems to show that when the velocities are well-constrained, the pattern speed and bulk velocities are similar.  But even in the same sources, the pattern speed can differ and stationary and high-velocity structures can be present.
  
• What's the source of X-ray hotspots? For luminous hotspots, the X-rays seem to come from Synchrotron Self-Compton emission, where the magnetic fields are in equipartition; hot spots in lower luminosity jets seem to shine due to synchrotron emission alone (at least in FRII sources; astro-ph/0405516).
• Why do BL Lacs look so different from other quasars (large variability, no broad emission lines)?  One theory is that they are beamed sources -- that we're looking right down the radio jet in these sources. Another theory is that they are gravitationally-lensed quasars, but some data seems to suggest that at least some BL Lacs are not gravitationally lensed (see astro-ph/0406284).
  

Classification of AGNs

Radio Quiet AGNs

Radio Loud AGNs

• Radio-loud AGN are distinguished by their relativistic jets, which apparently are launched from within a few tens of Schwarzschild radii of the central black hole (see Urry, 2003 for a nice review).
• Why are radio-loud galaxies radio-loud? This isn't well known. Some have claimed that it's at least weekly dependent on black hole mass (see, e.g., Laor 2000) but this has been disputed (Woo & Urry 2003 and Ho 2002). So, it seems that black hole mass isn't the only parameter dictating radio loudness; other possibilities include the black hole spin and accretion rate. Others have been theorizing that at least there seem to be lower and upper limits on how much radio power a certain black hole mass can produce, with the two limits split by about 5 orders of magnitude between bounds with L_{5 Ghz} varying as M_{BH}^2.5 (Dunlop 2003; Jarvis & McClure 2003)
• Like Radio-Quiet AGNs, there seem to be obscured radio-loud AGNs and unobscured radio-load AGNs. These are refered to as Narrow-Line Radio Galaxies (NLRG) and Broad Line Radio Galaxies (BLRG), respectively. The BLRGs are basically the low luminosity radio-loud quasars, so I'll group them in the Radio Loud Quasars section.

• Fanaroff-Riley Type 1 (FR I): FR I's are best known for their distorted (not well collimated) jets that don't have distinct termination points.  They have jets that appear disrupted, rather like distorted plumes, and appear as two-sided, apparently subsonic jets. It also seems that there is a fairly good separation in radio energy emitted by FR I's and FR2's: FR I's have L_178 MHz < 10^25 W Hz^-1. These are somewhat derivative definitions of the class, however; originally, the FR classification came from first defining R_FR as the ratio between the distance from the core of the highest radio surface brightness to the distance of the low-power radio width. If R_FR < 0.5, the source was classified as a FR I radio galaxy. More intuitively, FR I galaxies are those radio galaxies with their regions of low surface brightness farther from the core, or highest surface brightness closer to the core. And if it still doesn't make much sense, take a look at the two example images below, from Alan Bridle's beautiful image library. FR I galaxies are considered, in the Unification Scheme to be the counterparts of BL Lac Objects.  Unlike many other AGNs, they are not believed to have obscuring material (aka, "the torus") that changes the view of the core depending on inclination angle.  They also do not often show emission lines like other quasars, and so are believed to be relatively low luminosity AGNs.
  

 
3C 31, an example of a FR I-type radio galaxy. Note the very distorted radio jets on the left-hand image, in red, which shows 21cm radio emission taken with the VLA at 5.5 arcsecond resolution, with the Digitized Palomar Sky Survey in blue. (From Alan Bridle's Image Gallery.)

• Fanaroff-Riley Type 2 (FR II): Simply put, they have jets that appear smooth and un-distorted, and appear as one-sided, apparently supersonic jets. Also, FR II's have L_178 MHz > 10^25 W Hz^-1. In the original classification scheme of Fanaroff & Riley, FR II galaxies were those with R_FR > 0.5.  In the Unification Scheme, FR II galaxies are considered the counterparts of Radio Loud Galaxies.  Also, it is generally believed that FR II galaxies have optically thick material obscuring the source.
• It's not known why FR II galaxies appear different than FR I galaxies.  Perhaps the jets are different (so FR Is are intrinsically different from FR IIs) or perhaps the galaxies have different environments (so they are extrinsically different).  Some sources have been found (Gopal-Krishna & Wiita 2000) that seem to have an FR I jet morphology on one side, and an FR II-like jet on the other, which argues for extrinsic differences between the two types.
  

3C 175, an example of a FR II-type radio galaxy. The jets here are extraordinarily well-collimated, and only one of them is visible, presumably due to Doppler-shifting of the photons from the counter-jet. (From Alan Bridle's Image Gallery.)

Low Ionization Nuclear Emission Region Galaxies (LINERS)

Ultra-Luminous Infra-Red Galaxies (ULIRGS)

Broad Absorption Lines

Generally Accepted(?) Results

• 
  
• They are generally indistinguishable from other radio-quiet quasars in emission lines or continuum properties.
• They have weak soft X-ray flux, possibly due to high columns (N_H > 10^22 cm^-2) of gas blocking the central continuum.
• BALQSOs are only rarely radio-loud.
• Some of the absorption troughs begin in the middle of corresponding emission lines, and some are several thousand to tens of thousands of km s^-1 removed.
• Prominent lines: C IV, Si IV, N V, Ly alpha, O VI
• 10% to 20% of the time: Mg II, Al III are present.
• Column density measurements of absorption lines are complicated, as one has to consider partial covering, velocity dependences, and scattered light. All other calculations (much of the work before 1997) yield lower-limits to the columns given in BALQSOs.
• The covering fraction of material in BALQSOs is probably much less than unity, since we see much more absorption than emission, and high polarization in the absorption troughs. (However, if quasars do not emit isotropically, it is difficult to estimate the covering fraction from the fraction of BALQSOs that we see, as is typically done).

Recent Work

Radio Jets

Generally Accepted(?) Results

Recent Work

AGN Hosts/Environment & Fueling

Generally Accepted(?) Results

Recent Work

• Crenshaw,Kraemer, & Gabel (2003): NLS1's in the HST archives are much more likely to have large-scale stellar bars (at approximately beyond 1 kpc) than Broad-line Sy1's, and that the fraction of NLS1 with bars increases with decreasing H-beta FWHM.
• Maia, Machado, & Willmer (2003): In their SSRS2 survey, they find that 3% of their galaxies are Seyferts; within those Seyferts, the ratio of Sy2 to Sy1 galaxies is roughly 3 to 1. Seyferts seemed twice as likely to have bars as normal galaxies, were predominantly spirals of type Sb or earlier or in galaxies with a perturbed appearance, and were more likely to have a companion than normal galaxies.

AGN Evolution

Generally Accepted(?) Results

Recent Work