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

6.2. Metal Enrichment In The High Redshift Ly Forest

True to our self-imposed restraints we will deal here only with the low column density forest. For the reasons already mentioned (large transverse sizes, small velocity differences over hundreds of kpc, probable flattening, absence of voids, weak LOS correlation, low Doppler parameters, and the low column densities themselves) the typical Ly system at z ~ 3 is quite unlikely to be physically associated with a galaxy, other than being part of the gaseous matrix from which galaxies form. One puzzling result remains, however: the finding of widespread metal enrichment in the Ly forest.

Individual high column density systems with very low metallicities (of order 10^-3 ) exist (eg. Chaffee et 1985), but there is no absorption system known with column density above log N(HI) > 16 and a primordial composition. The possibility that there may be a transition at a certain column density from a metal enriched gas to a primordial gas has led to intensive searches for weak metal lines in low column density absorbers. Norris et al (1983), Williger et al (1989), and Lu (1991) used a "shift and stack" method to search for various metals in the low column density forest. To maximize the signal-to-noise ratio the spectra are shifted to the rest frame indicated by each Ly, and added, and the expected positions of metal lines are searched for a signal. A tentative detection of CIV, corresponding to a carbon depletion relative to solar of [C/H] - 3.1 was made by Lu et al 1991. Meyer & York (1987) pointed out, that data with an increasing S/N ratio show increasing numbers of individual weak CIV systems. The subject attracted renewed interest when Keck spectra showed that most Ly systems with HI column densities 10¹⁵ and roughly half of all Ly systems with column densities > 3 × 10¹⁴ cm^-2 have associated CIV lines, corresponding to a typical metallicity of Z ~ 10^-2Z (Cowie et al 1995; Tytler et al 1995; Songaila & Cowie 1996). Unfortunately, the detection threshold for CIV is close enough to make it hard to determine whether the decreasing rate of detections is a genuine turnover to primordial composition below a few × 10¹⁴, or just a selection effect. It is interesting that in the CDM models a column density contour of 10¹⁴ at redshift 3 delineates the transition between a continuous filamentary structure in the universe, with typical widths of less than 100 kpc, and the voids. Galaxies would have to spill metals only within the filaments to create the widespread metal-enrichment observed, and they could still have left most of the volume of the universe pristine. Thus a drop in metallicity at a few × 10¹⁴ would not come unexpected. A relatively uniform metal abundance across the whole column density range could be an indication of a earlier phase of nucleosynthesis.