|Annu. Rev. Astron. Astrophys. 1994. 32:
Copyright © 1994 by . All rights reserved
7.3. The Effect of Microlensing on Quasar Luminosity
Evidence for the microlensing of unmacrolensed quasars may come from studying their luminosity variations (Peacock 1986, Kayser et al 1986, Schneider & Weiss 1987, Refsdal & Starbell 1991, Lewis et al 1993), and there may already be cases of this. In particular, Nottale (1986) claims that lensing by low mass objects may explain some optically violently variable quasars. For example, the quasar 0846+51 brightened by 4 magnitudes in a month and then dimmed by 1 magnitude in a few days. The fact that its line of sight is only 12 arcsec from a galaxy suggests that the variation may result from microlensing by one of the halo objects, in which case the mass of the halo object must be in the range 10-4 to 10-2 M.
More dramatic, but no less controversial, evidence for the effect of microlensing on quasar luminosity comes from Hawkins (1993), who has been monitoring 300 quasars in the redshift range 1-3 over the past 17 years using a wide-field Schmidt camera. He finds quasi-sinusoidal variations of amplitude 0.5 m on a 5y timescale and he attributes this to lenses with mass ~ 10-3 M. The crucial point is that the timescale decreases with increasing z, which is the opposite to what one would expect for intrinsic variations (and these would be on a shorter timescale anyway). The timescale also increases with the luminosity of the quasar. He tries to explain this by noting that the luminosity should increase with the size of the accretion disk, but this only works if the disk is larger than the Einstein radius of the lens(about 0.01 pc), which is questionable. Another worrisome feature of Hawkins' claim (cf Schneider 1993) is that he requires the density of the lenses to be close to critical (so that the sources are being transited continuously). In this case, Big Bang nucleosynthesis constraints require the lenses to be nonbaryonic, so he is forced to invoke primordial black holes.