Department of Astronomy, University of Virginia, P.O. Box 400325, Charlottesville, VA 22904-4325

Abstract:

We present a new set of model predictions for 16 Lick absorption line indices from $H\delta$ through Fe5335, and UBV colors for single stellar populations with ages ranging between 1 and 15 Gyr, and [Fe/H] ranging from -1.3 to +0.3, and variable abundance ratios. The models are based on accurate stellar parameters for the Jones library stars and a new set of fitting functions describing the behavior of line indices as a function of effective temperature, surface gravity, and iron abundance. The abundances of several key elements in the library stars have been obtained from the literature in order to characterize the abundance pattern of the stellar library, thus allowing us to produce model predictions for any set of abundance ratios desired. We develop a method to estimate mean ages and abundances of iron, carbon, nitrogen, magnesium and calcium that explores the sensitivity of the various indices modeled to those parameters. The models are compared to high S/N data for Galactic clusters spanning the range of ages, metallicities and abundance pattern of interest. Essentially all line indices are matched when the known cluster parameters are adopted as input.

Comparing the models to high-quality data for galaxies in the nearby universe, we reproduce previous results regarding the enhancement of light elements and the spread in the mean luminosity-weighted ages of early-type galaxies. When the results from the analysis of blue and red indices are contrasted, we find good consistency in the [Fe/H] that is inferred from different Fe indices. Applying our method to estimate mean ages and abundances from stacked SDSS spectra of early-type galaxies brighter than $L^\star$, we find mean luminosity-weighed ages of the order of $\sim$ 8 Gyr and iron abundances slightly below solar. Abundance ratios, [X/Fe], tend to be higher than solar, and are positively correlated with galaxy luminosity. Of all elements, nitrogen is the more strongly correlated with galaxy luminosity, which seems to indicate secondary nitrogen enrichment. If that interpretation is correct, this result may impose a lower limit of 50-200 Myr to the timescale of star formation in early-type galaxies. Unlike clusters, galaxies show a systematic effect whereby higher-order, bluer, Balmer lines yield younger ages than $H\beta$. This age discrepancy is stronger for lower luminosity galaxies. We examine four possible scenarios to explain this trend. Contamination of the bluer indices by a metal-poor stellar population with a blue horizontal branch cannot account for the data. Blue stragglers and abundance-ratio effects cannot be ruled out, as they can potentially satisfy the data, even though this can only be achieved by resorting to extreme conditions, such as extremely high [O/Fe] or specific blue-straggler frequencies. The most likely explanation is the presence of small amounts of a young/intermediate-age stellar population component. We simulate this effect by producing two-component models and show that they provide a reasonably good match to the data when the mass fraction of the young component is typically a few %. If confirmed, this result implies star formation has been extended in early-type galaxies, and more so in less massive galaxies, which seems to lend support to the ``downsizing'' scenario. Moreover, it implies that stellar population synthesis models are capable of constraining not only the mean ages of stellar populations in galaxies, but also their age spread.

galaxies: abundances - galaxies: evolution - galaxies: elliptical and lenticular, cD - galaxies: stellar content - Galaxy: globular clusters - stars: fundamental parameters