Annu. Rev. Astron. Astrophys. 1998. 36: 267-316 Copyright © 1998 by Annual Reviews. All rights reserved

5. THE COSMOLOGICAL SIGNIFICANCE OF THE LYMAN ALPHA FOREST

Over the past few years semi-analytical work and in particular hydrodynamic simulations of hierarchical structure formation have gradually led to a minor Copernican shift in our perception of the material content of the high redshift universe. If the inferences (discussed below) are interpreted correctly the intergalactic medium is the main repository of baryons down to redshifts at least as low as z ~ 2. If so, then high redshift galaxies - in absorption line parlance "Lyman limit" or "damped Ly systems" - are mere tracers of the matter distribution. The simulations show that the Ly forest is produced by a hierarchy of gaseous structures, with typical shapes changing from sheets through filaments to spherical galactic gaseous halos, as the column density increases. Perhaps most importantly, Ly forest lines closely reflect gravitationally induced density fluctuations in the general matter distribution. Given the relatively simple physics of this baryonic reservoir and the enormous sensitivity of the observations Ly forest spectra should make excellent and largely unbiased probes of structure formation at high redshift.

THE LARGE BARYON CONTENT     The fraction of matter incorporated into galaxies or still left in the intergalactic medium depends strongly on the structure formation model. To calculate the baryon content of Ly clouds we need to know the ionization correction, as most of the gas is highly ionized. For a given ionizing radiation field the degree of ionization depends on the density and thus, for a given observed column density, on the spatial extent of the gas. Deriving the mass content then requires fixing the size (or scale height) of the clouds either from measurement, or from theoretical prejudices. For example, the fraction of mass required to cause the observed amount of absorption can be quite large for gravitationally confined, extended, baryonic clouds (Black 1981). In contrast, the small baryon content expected if the Ly were caused by pressure confined clouds (Sargent et al 1980) is largely a result of the small cloud sizes adopted. By using a suitable choice of parameters, the Ly forest can be made to contain anything from a negligible fraction up to virtually all of the baryons, and still be consistent with the observations (Meiksin & Madau 1993). Specifically for the CDM minihalo model, Petitjean et al (1993b) found that the Ly forest clouds had to contain most of the baryons at redshift 2-3, in order to match the observed column density distribution function. This is in agreement with Shapiro et al (1994), who found that in a CDM model the fraction of baryons not yet collapsed into galaxies should be on the order of 50-90% . Independent of the cosmological model, the large transverse sizes of Ly absorbers measured from QSO pairs give another, indirect indication that the baryon density in Ly clouds must be large, or the absorbers must be extremely flattened (Rauch & Haehnelt 1995).

A FOREST OF LINES, OR A FLUCTUATING GUNN-PETERSON EFFECT ?

Under the influence of gravity the intergalactic medium becomes clumpy and acquires peculiar motions, and so the Ly (or GP) optical depth should vary even at the lowest column densities (Black 1981; McGill 1990; Bi et al 1992; Miralda-Escudé & Rees 1993; Reisenegger & Miralda-Escudé 1995). In a CDM dominated structure formation scenario the accumulation of matter in overdense regions reduces the optical depth for Ly absorption considerably below the average in most of the volume of the universe, leading to what has been called the fluctuating Gunn-Peterson phenomenon. Traditional searches for the GP effect that try to measure the amount of matter between the absorption lines are no longer very meaningful as they are merely detecting absorption from matter leftover in the most underdense regions. If this is not taken into account the amount of ionizing radiation necessary to keep the neutral hydrogen GP absorption below current detection limits can easily be overestimated.

As another consequence, the distinction between the low column density Ly forest "lines", and the GP "trough", becomes somewhat artificial. Bi and collaborators (Bi et al 1992; Bi 1993; Bi & Davidsen 1997) have shown that the optical depth fluctuations corresponding to the linear regime of gravitational collapse in the intergalactic medium can give a remarkably realistic representation of the Ly forest (ignoring the higher column density lines, which are produced from non-linear structures, e.g., minihalo type objects). Their semi-analytical work is based on a log-normal density fluctuation field. For low densities where dissipation is not important the collapse of dark matter and baryons differs mainly by the presence of the gas pressure which effectively smooths the baryons distribution on scales below the Jeans length. Bi et al treated the pressure as a modification to the power spectrum of the baryon density contrast _b, suppressing power on scales below the Jeans length:

(16)

where _J is the Jeans length, k the wavenumber, and _DM the dark matter overdensity. This method can elucidate many of the basic features of low column density Ly clouds. The schematic treatment of the equation of state and the lack of inclusion of shock heating limit the approach, however, to overdensities of < 5, where gas physics beyond the Jeans criterion is not very important.