8.6. Anisotropic Emission and Beaming
Some bright quasars are known, to vary, by a large amplitude, over short periods of time. These are named OVVs (Optically Violent Variables). They are all radio sources that show some of the characteristics of the BL Lac objects, such as a non-thermal continuum and high degree of polarization. It is believed that relativistic beaming is important in this case, and the objects are seen close to the direction of the beam.
Recent observations show rapid emission line variability in some of these objects, on time scales shorter than estimated from the L - r relations of other AGNs. Since beaming is important, the ionizing flux is highly anisotropic, and line emission from clouds along the beam is much more intense than in other directions. The situation is very similar to the one illustrated in the top half of Fig. 19, where the beam illuminates only a small part of the shell-shaped BLR, and is the angle between the beam and the observers' direction. The time lag, r(l - cos ) / c, is short since is small.
There are several difficulties with this picture. First, for a double sided beam, there is line emission from the opposite side of the source. Unless it is obscured, this should arrive to the observer at a much later time. No such event has yet been observed. Second, the observed line profiles are usually smooth and symmetric. This is difficult to reconcile with a one-sided, jet-like system, whose center of gravity coincides with the central radiation source. Perhaps other mechanisms, such as radiation pressure (see next chapter), control the shape of the line profile. Nevertheless, beaming is a potentially important process that ought to be studied further since it may play a role in many objects whose ionizing beam is pointed away from us.
Observations: See chapter 2.
Theory: The first theoretical paper on the subject is by Bahcall, Kozlovsky and Salpeter (1972). More recent and detailed calculations are described in Blandford and McKee (1982). The application of the cross-correlation method to line variability studies is discussed in Gaskell and Sparke (1986) and Robbinson and Perez (1990). The uncertainties in the cross-correlation method are discussed by Gaskell and Peterson (1987), Maoz and Netzer (1989, the CCPD method) and Edelson and Krolik (1989, the Discrete Correlation function). Deconvolution and light-curve modeling are discussed in Maoz et al. (1991).