Annu. Rev. Astron. Astrophys. 1992. 30: 311-358 Copyright © 1992 by Annual Reviews. All rights reserved

5.4 Large Scale Structure and Intergalactic Medium

A great deal of evidence has accumulated in recent years that the universe is inhomogeneous on scales of up to 50-100 Mpc (e.g. Geller & Huchra 1989, Saunders et al. 1991). Such inhomogeneity will affect the propagation of light and lead to fluctuations in the convergence and shear of light bundles (Section 3.3). This has been studied in a number of papers which discuss distortions in the images of distant sources and changes in the apparent luminosities of the sources (e.g. Kantowski 1969, Dyer & Roeder 1973, Hammer 1985, Dyer 1986, Dyer & Oattes 1988, Schneider & Weiss 1988ab, Kasai et al. 1990, Jaroszynski et al. 1990, Babul & Lee 1991, Bartelmann & Schneider 1991, Tomita 1991).

The discovery of the population of faint blue galaxies now provides a potential tool to study these effects. Specifically, large-scale mass inhomogeneities will induce correlated elliptical distortions in images of background sources, which could be measured by mapping large numbers of sources over areas of the sky ~ 1 square degree in size (Kristian 1967, Blandford et al. 1991, Miralda-Escudé 1991c, Kaiser 1992). The magnitude of the correlated ellipticity should be in the range 1-3% (depending on the source redshift distribution) if the cold dark matter cosmology is correct, with much larger signals should mass trace light. The same technique will also reveal any coherent structures such as walls or voids. Although an early attempt to measure some of these effects only gave upper limits (Valdes et al. 1983), recent improvements in techniques make it worthwhile to repeat the observations.

In addition to relatively smooth large-scale inhomogeneities, a fraction of the mass in the universe may be in the form of compact dark objects on mass scales ranging from stars to galaxy clusters. These objects will cause distortions in the images of cosmologically distant sources on angular scales ~ _E (Press & Gunn 1973). Roughly, the optical depth to lensing is ~ _E, the density parameter corresponding to mass within the projected Einstein rings of the lenses. For compact objects in the mass range 10^-2 - 10⁵ M, a limit of somewhat less than unity may be set on _E merely by the fact that the line-to-continuum ratio is relatively constant among quasars (Canizares 1982). Better limits could be obtained from future observations of cosmologically distant supernovae (Schneider & Wagoner 1987, Linder et al. 1988, Rauch 1991). With masses larger than 10⁵ M, one would expect occasionally to see image distortions (Blandford & Jaroszynski 1981), image doubling (Nemiroff & Bistolas 1990, Nemiroff 1991), rings (Burke 1992), or phantom images in radio maps (Kassiola et al. 1991). Current observations limit the ₀ in the form of these objects to be 0.1. Future observations, especially using the VLBA, may be used to place significantly stronger limits.

Lensing by cosmic strings has been studied by some authors. A straight string will produce a characteristic lensing pattern consisting of approximately equal-separation image pairs (Vilenkin 1984, 1986, Gott 1985) with similar parity. A promising example of this was discovered (Cowie & Hu 1987), but it may be merely a cluster of binary galaxies (Hewitt et al. 1990, Hu 1990). The image of an extended source, such as a galaxy, can exhibit a discontinuity if the source happens to lie behind a string (Paczynski 1986c), though this signature may not be obvious if strings exhibit small scale kinky structure as now thought (Bennett & Bouchet 1988). Small string loops are expected to have lensing properties similar to those of galaxies (Hogan & Narayan 1984).