Annu. Rev. Astron. Astrophys. 1997. 35: 445-502
Copyright © 1997 by . All rights reserved

Next Contents Previous


Fundamental results have been obtained from variability studies of AGN, but many important questions are still unanswered. Substantial progress is expected in the near future.

Regarding the radio-quiet AGN, the simultaneity of the continuum flux variations at optical and UV wavelengths and the very short doubling time scales (one week to several months) observed in low-luminosity AGN are incompatible with models wherein the rapidly variable component of the optical-UV continuum is emitted through viscous effects in an accretion disk. Because the amplitude of the variations on time scales of weeks is more than a factor of 2 at 1400 Å, at least half of the blue bump luminosity is involved. The best current model for the rapidly variable continuum emission is reprocessed emission from a dense medium irradiated by the variable X rays.

For high-luminosity AGN, the importance of this irradiation process is uncertain because the relevant observations have not yet been carried out. The most interesting question - whether and how the continuum fraction emitted through reprocessing varies with absolute luminosity (accretion rate) and/or with the mass of the central black hole - remains unanswered. The organization of a series of quasisimultaneous optical-UV-X-ray observations is therefore a priority in order to advance in this field.

The rather good knowledge of the variability characteristics that we now have, at least for low-luminosity AGN, should stimulate new theoretical investigations on the origin of the continuum variability, both time scales and amplitudes. While turbulence and magnetic reconnection may be a sufficient explanation for the short time-scale X-ray variability, the mechanisms for longer term variations are essentially unknown.

Results crucial for our understanding of the inner region of radio-quiet AGN have been obtained from variability studies of the broad-line gas emission, including (a) a measurement of the dimension of the emitting region (dimension that depends on the line and on the velocity); (b) indications that the gas has roughly virialized motions in the gravity field of the central black hole, whose mass can then be derived from the observed motions; and (c) support for the accretion disk plus wind working model wherein the LIL come from the accretion disk, and the HIL are emitted by a wind (which could be magnetically or radiatively driven from the disk).

Among future investigations, we see three priorities. The first is to confirm (or refute) the accretion disk plus wind model, which at present seems to be the most promising model, by collecting extensive data sets combining high S/N and large number of epochs. Second, and possibly even more important, is to verify if the models developed for low-luminosity AGN are valid for high-luminosity AGN, the brightest objects in the universe. This can be done only if we collect data on high-luminosity AGN that are of a quality similar to the low-luminosity AGN data. This requires more powerful instruments and a consistent effort over much longer periods than for low-luminosity AGN. The third priority is to explore the entire parameter space defined by mass and mass accretion rate. This requires monitoring the continuum and line variations in very low-luminosity AGN (low accretion but unknown mass), such as the LINERS and the galaxy centers displaying faint UV sources or very weak broad lines. Similar observations should be carried out on NLS1, which might be the AGN with the highest accretion rate.

Closely related to these questions are the recent observations of broad, variable Fe K lines (Tanaka et al 1995, Yaqoob et al 1996) and of absorption features originating in a warm absorber (Iwazawa et al 1997). All these observations will add impetus to the active field of investigations of the structure and stability of AGN accretion disks, the role of the magnetic fields, and the magnetically or radiatively accelerated outflows.

As for the strongly nonthermal activity of blazars, it is now reasonably well understood in phenomenological terms. The basic model consists of a relativistic jet closely aligned with the line of sight, filled with energetic particles that emit synchrotron and Compton-scattered radiation spanning radio through gamma-ray wavelengths. The radiating particles must have a bulk relativistic motion, beaming the radiation toward us and making the blazar appear more luminous and more rapidly variable. Beyond this, models differ, notably in the structure and physical parameters of the jet and in the nature of the seed photons to be upscattered to the gamma-ray range.

The detection of short lags in the UV to X-ray range in Mrk 421 and PKS 2155-304 indicates that, in HBL at least, energy injection in the jet occurs in a top down scenario. Because the spectral variability of LBL objects in the IR to UV range appears similar to that of Mrk 421 and PKS 2155-304 in X rays, it is important to look for similar lags in LBL at IR to UV wavelengths. In principle, this could be done with intensive ground-based monitoring but would be much more likely to succeed with space instrumentation (HST and ISO) owing to the more extended wavelength range accessible and to the quasicontinuous coverage and even sampling. Comparison of flare evolution in wavelength and time in these very different sources, HBL and LBL, will give important clues to the differences in their jets.

Determining the nature of the seed photons (possibly different in different objects) is important because it will allow better determination of the physical conditions in the jet and eventually understanding of whether and how the environment influences the jet properties. Because model predictions for correlated variability between the synchrotron and Compton emission differ, multiwavelength campaigns offer the best means to discriminate. Substantial progress is expected over the next few years. Repeated multiwavelength campaigns on the brightest sources should enable discrimination among different sources of seed photons, especially with studies of multiwavelength variability on short time scales, during which the external photon field should not change substantially.

It is unfortunate that the available time for these programs is limited by the residual amount of gas for the EGRET experiment. One of the critical difficulties is that ground-based observations in the optical (as IUE is not available any longer) cannot be arranged quickly enough after the discovery of inherently unpredictable gamma-ray flares, given their rapid intraday time scales. In order to measure cross correlations and possible lags between gamma-ray and optical flares, optical-UV monitors on board the next generation gamma-ray experiments are needed.

Jet models imply that energy is transported in a dissipationless way from very small scales, within a few Schwarzschild radii of the central black hole to relatively large distances along the jet, a light day or more. At present we have no clear picture of how this happens. We can infer from X-ray and gamma-ray observations that the innermost region is unlikely to be filled with photons, as the resulting pair cascades would redistribute energy from gamma-ray to X-ray energies. Further work in this area is critical, as it holds the key to understanding how energy is extracted from the central black hole, the fundamental issue in the study of AGN.


It is a pleasure to thank our colleagues for comments and discussions at various stages of the preparation of this review, in particular R Blandford, S Collin-Souffrin, T Courvoisier, G Ghisellini, K Horne, B Hufnagel, L Lucy, G Madejski, S Phinney, E Pian, P Schneider, A Treves, R Terlevitch, and S Wagner.

Next Contents Previous