ACTIVE GALAXIES AND QUASISTELLAR OBJECTS,, EMISSION LINE REGIONS GREGORY A. SHIELDS Twenty-five thousand light-years away, behind the stars of Sagittarius, the nucleus of our galaxy lies hidden by intervening gas and dust. Peering through the dust with instruments sensitive to radio and infrared radiation, astronomers are finding there a bizarre scene. Clouds of ionized gas, heated by ultraviolet radiation of unknown origin, orbit a black hole with the mass of three million suns. Striking as it is, this activity is dwarfed by the powerful active galactic nuclei(AGN) observed in some other galaxies. Drawing their energy from a source no bigger than the solar system, the brightest AGN far outshine their host galaxy, with its hundreds of billions of stars. Their radiation, spanning all bands of the electromagnetic spectrum, consists of a "continuum"and "emission lines. "The continuum, with its energy smoothly distributed over all wavelengths, may be radiated by cosmic rays or by a hot disk of gas orbiting the black hole. The emission lines, occurring at discrete wavelengths in the infrared, optical, and ultraviolet, represent photons of light emitted by atoms in hot, ionized clouds of gas. The clouds' location, chemical composition, physical conditions, and motion can be derived by careful observation and theoretical analysis. Such information may help to answer the following basic questions: Do AGN really contain giant black holes? Does the energy derive from gas falling into these holes? What is the source of the gas? Galaxies containing an active nucleus are called active galaxies. These include Seyfert galaxies, BL Lac objects, radio galaxies, and QSOs. In 1943, Carl Seyfert described a class of spiral galaxies, otherwise normal, that have brilliant, starlike nuclei. The spectrum of a Seyfert nucleus shows a continuum whose energy distribution, polarization, and lack of absorption lines imply that it is not ordinary starlight. The continuum now is known to extend from radio and gamma ray frequencies. Superimposed on this "nonthermal" continuum are broad emission lines at the wavelengths familiar from nebulae in our galaxy but somewhat smeared out in wavelengths. QSOs are believed to be active galaxies in which the image of the host galaxy is lost in the glare of the brilliant nucleus. AGN occur over a hudge range in luminosity, from 10x4L* for the brightest QSOs to only 10x6 L* for the dimmest known Seyfert nucleus (fainter than the brightest normal stars), where L* denotes the luminosity of the Sun. The character of the continuum and emission lines of AGN is preserved over this entire range of luminosity. The emission lines of AGN fall into two categories according to the line profile (see Quasistellar Objects, Spectroscopic and Photometric Properties). The profile describes how the intensity of the line radiation varies with wavelength offset from the central wavelength. The light is spread out in wavelength because the motion of the gas in various parts of the emitting region causes a Doppler shift. The blue(short-wavelength) wing of the profile corresponds to photons emitted by gas moving toward the observer, and vice versa. The profile's breadth in wavelength gives the speed of the gas. Typically, AGN show a combination of "broad lines", with widths up to -10, 000km s-1 and "narrow lines" with widths -1000 km s'. These are attributed to two distinct regions in the nucleus. The broad-line region (BLR) involves fast-moving clouds of dense gas near the central engine, whereas the narrow-line region (NLR) involves more remote clouds with lower velocities and densities. Both regions are believed to be powered by the nonthermal continuum. THE NARROW-LINE REGION The relative intensities of the narrow lines resemble those of planetary nebulae. The intensity ratios of certain pairs of lines give the gas temperature and density (atoms per cm x3). Typical values for the NLR are T = 1-2x10x4 K and N = 10x3-10x6 cm-3. The elements are spread over a wide range of ionization states from Ox0 to fe+9. The most prominent are from H, He, C, N O, Mg, Si, S, and Fe, once or several times ionized. In planetary nebulae, the gas is heated and ionized by "photoionization, " in which ultraviolet photons from the central star eject electrons from atoms in the surrounding gas. A similar model explains the narrow lines of AGN, but here the ionizing radiation is the high-frequency extrapolation of the continuum observed in the optical and ultraviolet. Computer models of the physical processes in the photoionized clouds, called photoionization models, can explain the line intensities if the gas has a normal chemical composition. The NLR gas typically extends to distances of a few hundred light-years from the continuum source. The emitting clouds fill only a small fraction of this volume; and there is uncertainty about what if anything, fills the rest of the volume. The mass of gas in the NLR is -10x5M*, only a tiny fraction of the combined mass of the stars in the same region. Arguments involving the asymmetry of the line profiles (unequal brightness of the red and blue wings) suggest that gas is moving outward, but its origin is uncertain. Some observations indicate that within the NLR, denser gas moves at higher velocities. THE BROAD-LINE REGION The BLR is interesting because of its proximity to the center of the nucleus. Do its high velocities represent orbital motion around the black hole, infall toward it, or outflow? Does this gas actually fuel the energy source, or is it merely a by-product? Can its motion be used to weigh the black hole? The BLR gas is very dense by nebular standards. The absence of forbidden emission lines such as [OIII]^5007 implies N *10x8cm-3. On the other hand, the presence in the spectrum of intercombination lines such as [CIII]^1909 implies N*10x10 cm-3. Remarkably, the implied value of -10x9 cm-3 applies over the whole range of AGN luminosities. Gas at these densities radiates efficiently, so that only a few solar masses of gas are needed to produce the broad-line emission. The line intensities ca be fairly well explained by photoionization models. The ionization structure of the emitting clouds consists of a surface layer highly ionized by the ultraviolet continuum and deeper layer ionized by x-rays. This layer emits the Balmer lines and continuum of hydrogen and lines of OI, MgII, CaII, and Fe II. The complex physics of x-ray ionization and radiative transfer in optically thick lines requires elaborate computer programs. The continuum of AGN may be beamed in certain directions; and this contributes to the uncertainty as to whether or not the observed continuum, extrapolated to high frequencies, contains enough energy to power the BLR. Also uncertain is the fraction of the continuum intercepted by the BLR; some arguments suggest values -10x-1, with larger values in lower luminosity AGN. Photoionization models involve the "ionization parameter, "U=#i/Nc, where #i(photons cm-2s-1)is the flux of ionizing radiation striking a typical cloud and c is the speed of light. Larger values of U correspond to higher degrees of ionization in the gas, and values U-10x2 best reproduce the observed spectrum. This value varies little over the entire range of AGN luminosities, which in turn implies a uniform value of #i. The radius of the BLR, RB, and the central ionizing luminosity, Li, are related to #i through the inverse square law. This gives RB=1 ly for a typical QSO, and RB must vary roughly as Li1/2 over the range of AGN luminosities. The gas emitting the broad lines occupies a tiny fraction of the volume within RB. The traditional picture is a swarm of clouds, only -10x-4 ly thick, immersed in some hot medium whose pressure holds the clouds intact. An alternative measure of RB has been used in recent years. The continuum of some AGN varies in brightness on time scales of days to years. The emission lines should change the brightness accordingly, but with a delay corresponding to the time required for light to cross the BLR. Observations confirm that more luminous AGN have larger RB; but in given AGN, this method gives values of rB as much as 10 times smaller than the "U method" described previously. This discrepancy is a puzzle. The nature of the BLR's motion is unclear. The clouds are opaque at the wavelengths of some emission lines, and simple arguments then imply that the line profile should be asymmetrical in one way for outflow and in the reverse way for infall. Generally, the observed profiles are symmetrical. Another argument involves time variations. For outflow, the blue wing of the line should change before the red wing, and conversely for infall. Observed variations usually are symmetrical. These observations argue against pure infall or outflow but allow gas moving in elliptical or circular orbits. One picture is that the line emission comes, not from clouds, but from the surface of an "accretion disk" of gas extending from near the hole's horizon to radii > 1 ly. (The horizon, or boundary, of a black hole of mass 10x8 M* has a radius comparable with the earth's orbit around the Sun. ) The AGN derive their energy from the inner disk, but radiation intercepted by the outer disk could produce the lines. This model gives a unified explanation of the lines and continuum, but it has trouble fitting the line profiles in detail. In another model, the broad lines come from interstellar clouds that fall near the hole just as orbit near the sun. A more exotic theory suggests that the broad lines are emitted by the gaseous debris of stars ripped apart by gravity as they pass too close to the black hole ("tidal disruption"). If the broad-line clouds are orbiting a black hole, or falling into it, the mass of the hole is given by m=RBv2/G, where v is the clouds' velocity. The masses implied are reasonable for the values of RB derived from variability, 10x8-10x9M* for QSOs and 10x6 -10x7M* for Seyfert galaxies. Such values are consistent with the Eddington limit, which specifies, for a given luminosity, the minimum mass that can accrete gas by overpowering the outward force of radiation pressure. Nearby galactic nuclei, such as m31, show evidence for a black hole of mass -10x7 M*; but in these cases, there is no AGN energy source. CONCLUSIONS The emission lines of AGN currently provide one of our few clues to the conditions prevailing at distances -10x-1 - 10x3 ly from the central object. The lines hold the promise of measuring the mass of the black hole and determining the nature of its fuel supply. Gradually, we are unraveling the secrets of the emission lines, but today there are more questions than answers. Additional Reading Burbidge, G. and Burbidge, E. M. (1967). Quasi-Stellar Objects. W. H. Freeman, San Francisco. Davidson, K. and Netzer, H. (1979). The emission lines of quasars and similar objects. Rev. Mod. Phys. 51 715. Miller, J. S. and Osterbrock, D. E. eds. (1984). Active Galactic Nuclei and Quasi-Stellar objects. University Science Books, San Francisco Osterbrock, D. E. (1989). Astrophysics of Gaseous Nebulae and active Galactic Nuclei. University Science Books, Mill Valley, Calif. Shaffer, D. B. and Shields, G. A. (1980). Why all the fuss about quasars? Astronomy Magazine 8 (No. 10)6. Strittmatter, P. A. and Williams, R. E. (1976). The line spectra of quasi- stellar objects. Ann. Rev. Astron. Ap. 14 307. Weedman, D. W. (1986). Quasar Astronomy. Cambridge University press, Cambridge. See also Active Galaxies, Seyfert Type; Galactic Center; Quasistellar Objects, Spectroscopic and Photometric Properties.