If the different colour subpopulations of GCs are formed in different events then they may contain clues to the formation history of their parent galaxies. For example a two-peak colour distribution may result, if in addition to a primary initial collapse population of GCs, a secondary population of GCs were formed either in a merger-induced starburst (Schweizer 1993; Ashman & Zepf 1992; Fritze - v. Alvensleben & Gerhard 1994; Fritze - v. Alvensleben & Burkert 1995) or else in some distinct secondary phase of cluster formation within the original galaxy (Forbes et al. 1997). Likewise the broad or multi-peaked colour distribution often observed in GC systems around cD galaxies may point to a series of GC formation events during the hierarchical assembly of the parent galaxy or to some protracted GC formation or accretion mechanism.
A well-known difficulty with the interpretation of colour distributions is the degeneracy of colours with respect to age and metallicity. While for Washington photometry there are well established and reliable calibrations of colours in terms of metallicity, the situation with HST broad band observations of GC systems is less clear. A better understanding of the formation of composite GC systems would be possible if separate age and metallicity distributions could be disentangled from an observed colour distribution.
A second issue concerns the interpretation of colours for young star cluster systems detected with HST in many interacting and starburst galaxies. The question is, if these YSC systems - at least some fraction of them - are the progenitors of GC systems. In an attempt to answer this question star clusters are being imaged with HST in an age sequence of interacting galaxies - from early stages of interaction through merger remnants up to E/S0s (e.g. Schweizer et al. 1996; Whitmore et al. 1995; Miller et al. 1997). With 10 m class telescopes, spectroscopy of the brighter members of young star cluster populations is becoming possible (Kissler-Patig et al. 1998; Brodie et al. 1998, but see also 4 - 5 m class spectra by Schweizer & Seitzer 1993 or Zepf et al. 1995). Spectroscopy will only be possible for a subsample of YSCs. Thus the determination of ages and metallicities from broad band colors will still be necessary.
It is thus desirable to study the evolution of broad band colours and absorption indices for single burst stellar populations of various metallicities using the most recent and complete stellar evolutionary tracks as well as careful colour and index calibrations.This allows one to obtain theoretical calibrations of broad band colours and indices in terms of metallicity over the full range of ages under investigation, i.e. from 107 yr to a Hubble time.
Since theoretical calculations for the evolution of stars are only available for a discrete grid of masses, some means for obtaining a smooth evolution of the composite population is needed. Applying the tracks as they are would create discontinuities because all stars of a given mass would reach the giant branch at the same time, dominating the integrated light until they die. This effect is large for populations with stars that have about the same age. The effect also increases with age of the whole population, since the differences in both the lifetimes and luminosities between the main sequence and the later stages increase with decreasing mass.
For this work, we use the Monte Carlo method to bypass this problem while still avoiding the interpolation of tracks with its accompanying danger of creating artificial states. This is described in detail in Sect. 2.3.
The star formation history of any stellar system can be described by a superposition of SSP models of different ages and metallicities. An example of this is given by Cellone & Forte et al. (1996) in their study of Low Surface Brightness galaxies or Contardo et al. (1998) who investigate the formation and evolution of galaxies in a cosmological scenario.
Copyright The European Southern Observatory (ESO)