paper

An optical/NIR survey of globular clusters in early-type galaxies III. On the colour bimodality of GC systems

arXiv:1201.3649 · doi:10.1051/0004-6361/201117169

Abstract

The interpretation that bimodal colour distributions of globular clusters (GCs) reflect bimodal metallicity distributions has been challenged. Non-linearities in the colour to metallicity conversions caused by the horizontal branch (HB) stars may be responsible for transforming a unimodal metallicity distribution into a bimodal (optical) colour distribution. We study optical/near-infrared (NIR) colour distributions of the GC systems in 14 E/S0 galaxies. We test whether the bimodal feature, generally present in optical colour distributions, remains in the optical/NIR ones. The latter colour combination is a better metallicity proxy than the former. We use KMM and GMM tests to quantify the probability that different colour distributions are better described by a bimodal, as opposed to a unimodal distribution. We find that double-peaked colour distributions are more commonly seen in optical than in optical/NIR colours. For some of the galaxies where the optical (g-z) distribution is clearly bimodal, the (g-K) and (z-K) distributions are better described by a unimodal distribution. The two most cluster-rich galaxies in our sample, NGC4486 and NGC4649, show some interesting differences. The (g-K) distribution of NGC4649 is better described by a bimodal distribution, while this is true for the (g-K) distribution of NGC4486 GCs only if restricted to a brighter sub-sample with small K-band errors (< 0.05 mag). Formally, the K-band photometric errors cannot be responsible for blurring bimodal metallicity distributions to unimodal (g-K) colour distributions. However, simulations including the extra scatter in the colour-colour diagrams (not fully accounted for in the photometric errors) show that such scatter may contribute to the disappearance of bimodality in (g-K) for the full NGC4486 sample. For the less cluster-rich galaxies results are inconclusive due to poorer statistics. [Abridged]

A&A accepted, 15 pages, 10 figures, 4 tables

References in corpus (12)

Cited by in corpus (42)