Absorption line spectroscopy of foreground gas in the spectra of background quasars has revealed some clear cases where neutral gas is present and associated with dwarf galaxies. Spectroscopy of Lyα and low-ionization metal lines can be combined to derive neutral gas phase metallicities. The damped Lyα absorbers (DLAs) in quasar spectra are the clearest cases of absorption by predominantly neutral gas regions. Here we present some results on neutral gas phase metallicities for cases where the DLA is clearly associated with a dwarf galaxy. We find that the neutral gas phase metallicities in these systems are similar to those in other DLAs. We argue that there may be many unrecognized cases where a DLA is actually associated with a dwarf galaxy even though there is a luminous galaxy within 100 kpc of the quasar sightline.

We summarize results from our recent SDSS survey for CaII quasar absorbers. The survey finds 435 absorbers at z<1.34, which corresponds to ~9 Gyrs of the Universe's history. Two CaII absorber populations are identified, which we call weak and strong. Metal abundance ratios derived from normalized composite spectra show that weak absorbers can be identified with halo-like gas, while strong absorbers are a mix of halo+disk-like gas. Consistent with these abundance ratio results, fluxed composites show that strong absorbers are ~6 times more reddened than weak absorbers (i.e., they contain more dust). Four individual examples of galaxies associated with CaII absorption are found in the SDSS images, and three are star forming galaxies (SFGs). Image composites show that galaxies associated with strong absorbers have mean luminosity-weighted impact parameters, b, which are ~2.4 times smaller than the weak absorbers (bstrong ~ 19 kpc versus bweak 45 kpc). Studies of CaII absorbers suggest links to molecular clouds, SFGs, and circumgalactic gas (CGG).

The damped Lyman-α absorption-line systems (DLAs) that are observed in quasar spectra arise in neutral-gas-rich regions of intervening galaxies. With the highest neutral hydrogen column densities observed (N(HI) ⩾ 2 × 1020 atoms cm−2), they are known to trace the bulk of the neutral gas content of the Universe, and are thus powerful probes of galaxy formation and evolution. However, DLAs are extremely rare, and since the Lyman-α line falls in the UV for redshifts z < 1.65, not many are known at low redshift due to the limited availability of space data. Our HST surveys for DLAs in strong MgII absorbers have been successful at showing that MgII can be used as an unbiased tracer of DLAs. We present new results on their incidence, or redshift number density, dn/dz, and cosmological neutral gas mass density, ΩDLA, at redshifts 0.11 ⩽ z ⩽ 1.65, and incorporate results from higher and lower redshift studies in the literature to derive the evolution of neutral gas in the Universe.

We present results on the properties and evolution of moderately ionized gas in the Universe based on intervening CIV absorption-line systems. We have compiled a database of CIV systems from the SDSS, and we find evidence for evolution in the incidence of CIV absorbers.

We give references to some of our work on the properties and evolution of the neutral gas component of the Universe (see reference list). The bulk of the observed neutral gas has been detected by identifying intervening damped Lyα (DLA) quasar absorption-line systems with N(H) ≥2 × 1020 atoms cm−2. We also present some initial results from a program to identify DLA absorbers near redshift z = 0.5 using Hubble Space Telescope ACS prism spectra (see Figure 1).

The spectra of the Sloan Digital Sky Survey (SDSS) are being used to construct a catalogue of QSO absorption lines, for use in studies of abundances, relevant radiation fields, number counts as a function of redshift, and other matters, including the evolution of these parameters. The catalogue includes intervening, associated, and BAL absorbers, in order to allow a clearer definition of the relationships between these three classes. We describe the motivation for and the data products of the project to build the SDSS QSO absorption line catalogue.

Using a large sample of MgII absorbers with $0.4<z<2.2$ detected by Nestor et al. (2005) in the Early Data Release of the SDSS, we present new constraints on the physical properties of these systems based on two statistical analyses: (i) by computing the ratio between the composite spectra of quasars with and without absorbers, we measure the reddening effects induced by these intervening systems; and (ii) by stacking SDSS images centred on quasars with strong MgII absorption lines and isolating the excess light around the PSF, we measure the mean luminosity and colours of the absorbing galaxies. This statistical approach does not require any spectroscopic follow up and allows us to constrain the photometric properties of absorber systems.

Sloan Digital Sky Survey (SDSS) quasar spectroscopy is yielding a database of strong low-ionisation MgII absorbers over the redshift interval $0.36<z<2.28$ which is over two orders of magnitude larger than anything previously assembled. Hubble Space Telescope (HST) UV spectroscopy has been used to measure neutral hydrogen column densities for a small subset of them. These data empirically show that MgII absorbers with rest equivalent widths $W^{\lambda2796}_0 \ge 0.6$ Å have a mean neutral hydrogen column density that is roughly constant at $N(HI) \approx 4\times10^{20}$ atoms cm$^{-2}$, with individual systems lying in the damped Ly$\alpha$ (DLA) and sub-DLA regimes. Since the MgII doublets generally exhibit saturation, the $W^{\lambda2796}_0$ values are an indication of the absorbers' velocity spreads. Thus, we can study neutral-gas-phase metallicities as a function of kinematics by forming SDSS composite spectra and measuring weak unsaturated metal lines that form in neutral gas (e.g., CrII, FeII, MnII, SiII, ZnII) as a function of $W^{\lambda2796}_0$. We use this method on SDSS composite spectra to show how metallicity and kinematics are positively correlated for large-$N(HI)$ absorbers, including trends related to dust depletion and the enhancement of $\alpha$-elements. We also discuss the need to account for selection effects in DLA surveys, and we make inferences about models for DLA absorption and their contribution to cosmic star formation.

We consider multiwavelength observations of Broad Absorption Line (BAL) QSOs which include optical spectrophotometry supplemented by HST-FOS UV spectrophotometry and/or ROSAT-PSPC x-ray observations. For moderate to high redshift objects, increasing the wavelength coverage into the UV permits observation of BALs due to different ions of the same element for several different elements for the first time. By employing appropriate assumptions, ionic column densities as a function of outflow velocity can be derived from these observations, resulting in constraints on the level of ionization and chemical composition of the BAL region (BALR) gas and the photoionization model itself.

The APM QSO survey is a quantitative survey aimed at finding a large sample (∼ 1000) of QSOs using broadly-based selection criteria applied to machine-scanned UK Schmidt Telescope direct and objective-prism plates. The survey is currently entering its third year and, as of August 1988, the sample consists of ∼ 700 QSOs with mJ ≥ 18.75 in the range 0.2 ≤ z ≤ 3.3. Preliminary analysis suggests that the sample is relatively free of the selection effects endemic to most QSO surveys based on slitless spectroscopy.

Observational constraints on models for broad absorption line (BAL) QSOs are discussed. The picture which emerges is one in which the broad absorption line region (BALR) contributes to the broad emission line region (ELR) as a high ionization component. Depending on QSO luminosity, the limits on the distance of the BALR from the central source are 1–1000 pc. A highly metal enriched and disk-like geometry for the region is indicated. The similarity between the absorption in some radio quiet BAL QSOs and the associated complexes of absorption in some radio loud QSOs may indicate that the BALR outflow can affect the region surrounding a QSO out to distances in excess of several hundred kpc.