Testing stellar population synthesis models with Sloan Digital Sky Survey colors of M31's globular clusters

TL;DR

This study compares SDSS colors of M31 globular clusters with stellar population models, revealing a significant color offset that highlights the need for improved models, especially in the g-r band.

Contribution

It provides the first large-scale comparison of SSP models with SDSS ugriz colors of M31 globular clusters, identifying a key discrepancy and testing updated models.

Findings

SSP models predict g-r colors too red by ~0.1 compared to observations.

The updated Maraston & Stromback model better matches observed colors.

The g-r offset is unlikely due to young stellar populations.

Abstract

Accurate stellar population synthesis models are vital in understanding the properties and formation histories of galaxies. In order to calibrate and test the reliability of these models, they are often compared with observations of star clusters. However, relatively little work has compared these models in the ugriz filters, despite the recent widespread use of this filter set. In this paper, we compare the integrated colors of globular clusters in the Sloan Digital Sky Survey (SDSS) with those predicted from commonly used simple stellar population (SSP) models. The colors are based on SDSS observations of M31's clusters and provide the largest population of star clusters with accurate photometry available from the survey. As such, it is a unique sample with which to compare SSP models with SDSS observations. From this work, we identify a significant offset between the SSP models and the clusters' g-r colors, with the models predicting colors which are too red by g-r\sim0.1. This finding is consistent with previous observations of luminous red galaxies in the SDSS, which show a similar discrepancy. The identification of this offset in globular clusters suggests that it is very unlikely to be due to a minority population of young stars. The recently updated SSP model of Maraston & Stromback better represents the observed g-r colors. This model is based on the empirical MILES stellar library, rather than theoretical libraries, suggesting an explanation for the g-r discrepancy.