EDP Sciences
Free access
Volume 414, Number 2, February I 2004
Page(s) 515 - 530
Section Stellar clusters and associations
DOI http://dx.doi.org/10.1051/0004-6361:20031639

A&A 414, 515-530 (2004)
DOI: 10.1051/0004-6361:20031639

Secondary Globular Cluster populations

U. Fritze-v. Alvensleben

Universitäts-Sternwarte, Geismarlandstr. 11, 37083 Göttingen, Germany
(Received 20 March 2003 / Accepted 20 October 2003)

This study is motivated by two facts: 1. The formation of populous star cluster systems is widely observed to accompany violent star formation episodes in gas-rich galaxies as e.g. those triggered by strong interactions or merging. 2. The Globular Cluster (GC) systems of most but not all early-type galaxies show bimodal optical color distributions with fairly universal blue peaks and somewhat variable red peak colors, yet their Luminosity Functions (LFs) look like simple Gaussians with apparently universal turn-over magnitudes that are used for distance measurements and the determination of ${\rm H_{\rm o}}$. Based on a new set of evolutionary synthesis models for Simple (= single burst) Stellar Populations (SSPs) of various metallicities using the latest Padova isochrones I study the color and luminosity evolution of GC populations over the wavelength range from  U through  K, providing an extensive grid of models for comparison with observations. I assume the intrinsic widths of the color distributions and LFs to be constant in time at the values observed today for the Milky Way or M 31 halo GC populations. Taking the color distributions and LFs of the Milky Way or M 31 halo GC population as a reference for old metal-poor GC populations in general, I study for which combinations of age and metallicity a secondary GC population formed in some violent star formation event in the history of its parent galaxy may or may not be detected in the observed GC color distributions. I also investigate the effect of these secondary GCs on the LFs of the total GC system. Significant differences are found among the diagnostic efficiencies in various wavelength regions. In particular, we predict the NIR to be able to reveal the presence of GC subpopulations with different age - metallicity combinations that may perfectly hide within one inconspicuous optical color peak. If the entire manifold of possible age - metallicity combinations is admitted for a secondary GC population, we find several cases where the resulting LF of the whole GC system is significantly affected and its turn-over could not serve as a reliable distance indicator. If, on the other hand, we assume some age - metallicity relation for GC populations, the second peak of the LFs vanishes and models indicate single-peak GC LFs even in GC systems with bimodal color distributions. A broad but sufficient age - metallicity relation is, for example, obtained if the secondary GC populations form in mergers of various spiral galaxy types from the ISM pre-enriched over the redshift range from $z \ga 4.4$ to $z \ga 0$. As a first illustrative example we apply our models to  V- and I-band data presented by Larsen et al. (2001) for blue and red peak GCs in three early-type galaxies. We point out the importance of having multi-band information to independently constrain ages and metallicities of different GC subpopulations and again stress the diagnostic potential of K-band data in addition to optical observations. The models presented here will be used for the interpretation of multi-wavelength data on GC systems in galaxies of various types, luminosities and environments as well as on young star cluster systems in interacting galaxies and mergers. By independently constraining ages and metallicities of individual clusters we expect to shed light on both cluster and galaxy formation scenarios.

Key words: galaxies: star clusters -- globular clusters: general -- galaxies: interactions -- galaxies: starbursts

© ESO 2004