1999A&A...351..841M

We analyse predictions for high-z damped Lymanα (dLyα) QSO absorber systems from a special type of disk galaxy evolution models. The models include secular gas infall onto the disk and viscous radial gas flows within the disk. In a previous study, we have shown that these models provide a reasonable fit for all relevant observational properties of the Galactic disk. We discuss the evolution with redshift of the frequency distribution of the HI column density, the metallicity, and the star formation rate. The model predictions are compared with observational data from dLyα QSO absorbers on the assumption that these absorber systems represent early evolutionary stages of present-day spiral galaxies. We study two groups of models with different infall histories. Slow disk evolution models with strong gas infall are in poor agreement with all observational constraints. Moderate infall models provide a better agreement. Although some of the observed properties of dLyα systems are well reproduced by the models, there remain serious discrepancies. We confirm earlier findings that the model predictions for observed properties of the absorber systems strongly depend on the role of dust at high redshift. We apply Monte Carlo simulations to take into account a bias against high column density/high metallicity systems due to extinction of the QSO light by dust in the absorber galaxy. The results of the simulations provide an improved fit of the observed redshift-metallicity relation after the bias is taken into account. However, the predicted abundance scatter for models of the Galactic disks appears too small at intermediate and low redshifts. Furthermore, a deficiency of high column density systems would be expected for z≲1.5 due to the extinction bias. A better agreement with observations is expected if dLyα systems are associated with various galaxy types. The observed HI column density distribution at low redshift must obviously be explained by the contribution of galaxies having high HI column densities but low metallicities. The observed abundance ratio of alpha elements to iron peak elements seems to be nearly solar after correction for dust depletion, whereas the models clearly predict an overabundance of alpha elements at high redshift. The models predict too a high number density of absorber galaxies at z=0 when the distribution of absorbers at high redshift is matched. We discuss possible solutions for these discrepancies.