2016A&A...592A..55V

Stellar halos and globular cluster (GC) systems contain valuable information regarding the assembly history of their host galaxies. Motivated by the detection of a significant rotation signal in the outer halo GC system of M 31, we investigate the likelihood of detecting such a rotation signal in projection, using cosmological simulations. To this end we select subsets of tagged particles in the halos of the Aquarius simulations to represent mock GC systems, and analyse their kinematics. We find that GC systems can exhibit a non-negligible rotation signal provided the associated stellar halo also has a net angular momentum. The ability to detect this rotation signal is highly dependent on the viewing perspective, and the probability of seeing a signal larger than that measured in M 31 ranges from 10% to 90% for the different halos in the Aquarius suite. High values are found from a perspective such that the projected angular momentum of the GC system is within ≤40° of the rotation axis determined via the projected positions and line-of-sight velocities of the GCs. Furthermore, the true 3D angular momentum of the outer stellar halo is relatively well aligned, within 35°, with that of the mock GC systems. We argue that the net angular momentum in the mock GC systems arises naturally when the majority of the material is accreted from a preferred direction, namely along the dominant dark matter filament of the large-scale structure that the halos are embedded in. This, together with the favourable edge-on view of M 31's disk suggests that it is not a coincidence that a large rotation signal has been measured for its outer halo GC system.