Fortunately, the field of active galactic nuclei is by no means stagnant. Interesting new results continue to be generated at a significant pace, some of them, inevitably, requiring re-evaluation of what I said in 1998 when the text of the book was completed. On this page I'll give brief comments on recent work that I find to be of the greatest interest. Rather than presenting fully-digested pedagogical accounts (there would be no way to keep up if I attempted to do that), I'll merely give brief summaries and links to on-line versions of the most significant papers.
January 7, 2000
1.) Are radio-loud AGN really restricted to elliptical galaxies, and radio-loud AGN to spirals?
As discussed in §13.2.3, nearby radio-quiet AGN are almost exclusively found in spiral galaxies, whereas (intrinsically more luminous) radio-loud AGN are segregated just as thoroughly into elliptical host galaxies. On this basis, many people had long thought that this division of labor extended to all AGN across the board, of every luminosity. However, recent HST imaging of quasars at modest redshift ( roughly) suggests that all quasars, both radio-quiet and radio-loud can have elliptical hosts (Boyce et al. 1998; McLure et al. 1999). Because even HST data of quasar hosts is a bit rough, this conclusion is based not on direct measurement of the morphology, but rather on the optical color of the integrated stellar light from the host, and on comparing the azimuthally-averaged radial distribution of surface brightness in these hosts to templates: the surface brightness of normal spiral galaxies declines exponentially with radius, whereas in ellipticals, the variation is more nearly , the ``de Vaucouleurs Law". At higher red-shifts, only hosts of radio-loud quasars have, as yet been studied; there, too, they appear to be found in elliptical galaxies (Lehnert et al. 1999).
2.) Beyond the Novikov-Thorne model
The simplest possible model for an accretion disk is one in which the disk is time-steady and axi-symmetric. When Novikov and Thorne first worked out predictions for the structure of disks near enough black holes for relativity to be important, they naturally adopted these simplifying assumptions. They also made a further assumption-that the - stress went to zero at the marginally stable orbit. In support of this assumption they argued that matter in the plunging region inside would fall so quickly into the black hole that at any given time the inertia of matter in that region would be far too small to exert any significant force on the disk proper.
There is now reason to believe that all three of these assumptions are wrong. Let us begin the argument by taking for granted that the magneto-rotational instability (§7.2.3) creates enough MHD turbulence to account for all the disk's angular momentum transport. The linear growth rate of this instability is no smaller in the relativistic region than in the Newtonian part of the disk, and there is no reason to suppose that the dissipation rate changes drastically near the marginally stable orbit, so we have every reason to believe that the level of MHD turbulence near is just as large (relative to the disk pressure) as it is farther out. If this is so, we may fairly ask, why should fall to zero as ? For this to happen would require some special change in the correlational properties of the turbulence.
Nonetheless, just for the sake of argument, let us suppose that this happens; that is, as , but the ratio of magnetic energy density to total pressure remains constant. As matter passes through , it accelerates rapidly inward. Because its electrical conductivity remains very high, the plasma remains very closely tied to the magnetic field, and any field loops with footpoints in gas remaining in the disk are drastically stretched. In fact, if there is no dissipation or reconnection to break them off (and the infall is so fast that it might be difficult for significant dissipation to occur), the energy density in magnetic field automatically becomes comparable to the kinetic energy density of the plasma as soon as the radial component of the velocity becomes comparable to the azimuthal component (Krolik 1999)! When this is the case, the magnetic stress at must be significant.
That time-steadiness and axi-symmetry are also poor assumptions likewise follows from this picture in which MHD turbulence drives accretion. After all, turbulence is hardly consistent with either assumption. In fact, detailed 3-d MHD simulations (using Newtonian dynamics in a Paczynski-Wiita potential, rather genuine relativistic dynamics) are now beginning to show that the fluctuations in almost every quantity in the inner part of the disk are order unity (Hawley 1999).
In the long-run, only well-resolved 3-d MHD simulations set in a Kerr metric and employing a reasonably realistic equation of state will be reliable guides to the true nature of disk dynamics. However, these preliminary studies are already starting to cast some doubt on the simplest picture. At the same time, they also raise some hope for solving several important observational puzzles. Their most immediate qualitative consequence is that angular momentum transport by MHD turbulence leads to non-steady accretion, and possibly with non-steady accretion efficiency as well. Here may lie the origin of the ubiquitous variability of black hole accretion sources.
These mechanisms may also point to a new device for tapping the spin-energy of a rotating black hole. The Blandford-Znajek effect (§5.2.3) uses black hole rotation to create electrical currents, which are then dissipated both at large distance and in the black hole's event horizon. By contrast, in the picture described in the previous paragraphs, matter in the plunging region is forced into rapid rotation by the strong frame-dragging just outside a spinning black hole's event horizon. This matter then does work, via magnetic connections, on more slowly-rotating gas farther out. By this means, energy is conveyed from the black hole spin to the body of the disk, where the long infall time gives plenty of time for it to be dissipated.