ar X iv : a st ro - p h / 99 07 42 8 v 1 2 9 Ju l 1 99 9 Ionized Gas in Damped Lyman - α Systems and Its Effects on Elemental Abundance Studies

Recent high-resolution observations of metal absorption lines in high-redshift damped Lyα systems have shown that Al III, a tracer of moderately-ionized gas, very often has a velocity structure indistinguishable from that of low-ionization gas. Regions of ionized and neutral hydrogen in these systems are likely cospatial. The higher-ionization Si IV and C IV absorption shows a much weaker or non-existent correlation with the low ionization material, implying that the regions traced by Al III are photoionized by a soft (stellar) spectrum, by a hard (power law) spectrum with a very low ionization parameter, or a combination of both. We discuss the ionization of the damped Lyα systems and use photoionization equilibrium models to make quantitative estimates of its effects on abundance studies in these systems. We show that ionization effects may be large enough to account for the observed dispersion in absolute metal abundances in damped Lyα systems, causing systematically higher abundances in lower column density systems. The observed Si/Fe and Zn/Cr ratios may systematically overestimate the intrinsic Si/Fe and Zn/Cr ratios, respectively, if ionized gas is present in these systems, thereby mimicking the effects of α-element enrichment or dust depletion. Subject headings: galaxies: abundances – intergalactic medium – quasars: absorption lines – radiative transfer 1. EVIDENCE FOR PHOTOIONIZED GAS IN DAMPED Lyα SYSTEMS Observations of damped Lyα systems (DLAs), which may represent the progenitors to modern disk galaxies, along the sightlines to high-redshift QSOs allow astronomers to trace the evolution of elemental abundances over 90% of the age of the Universe. This is typically achieved by comparing the measured column density of a single ion of a given element, X , with that of neutral hydrogen, H. The assumption is then made that the unobserved ionization stages make negligible contributions and N(X )/N(H) ≈ N(X)/N(H). The total H I column densities of DLAs are, by definition, N(H I) ∼ > 2 × 10 cm; if collected in a single cloud, such a large column density would imply small ionization corrections. Although some of these systems include such monolithic absorbers, high-resolution data show DLAs are often made up of a collection of several (in many cases ∼ > 5 − 10) lower column density clouds (Lu et al. 1996; Pettini et al. 1999; Prochaska & Wolfe 1999). Furthermore, both the Lu et al. (1996) and Prochaska & Wolfe (1999)