Accreted versus in situ Milky Way globular clusters

A large, perhaps dominant fraction, of the Milky Way's halo is thought to be built up by the accretion of smaller galaxies and their globular cluster (GC) systems. Here, we examine the Milky Way's GC system to estimate the fraction of accreted versus in situ formed GCs. We first assemble a high-quality data base of ages and metallicities for 93 Milky Way GCs from literature deep colour-magnitude data. The age-metallicity relation for Milky Way's GCs reveals two distinct tracks - one with near constant old age of similar to 12.8 Gyr and the other branches to younger ages. We find that the latter young track is dominated by GCs associated with the Sagittarius and Canis Major dwarf galaxies. Despite being overly simplistic, its age-metallicity relation can be well represented by a simple closed box model with continuous star formation. The inferred chemical enrichment history is similar to that of the Large Magellanic Cloud, but is more enriched, at a given age, compared to the Small Magellanic Cloud. After excluding Sagittarius and Canis Major GCs, several young track GCs remain. Their horizontal branch morphologies are often red and hence classified as young halo objects; however, they do not tend to reveal extended horizontal branches (a possible signature of an accreted remnant nucleus). Retrograde orbit GCs (a key signature of accretion) are commonly found in the young track. We also examine GCs that lie close to the Fornax-Leo-Sculptor great circle defined by several satellite galaxies. We find that several GCs are consistent with the young track and we speculate that they may have been accreted along with their host dwarf galaxy, whose nucleus may survive as a GC. Finally, we suggest that 27-47 GCs (about 1/4 of the entire system), from six to eight dwarf galaxies, were accreted to build the Milky Way GC system we see today.