Rotational Kinematics in the Globular Cluster System of M31: Insights from Bayesian Inference

TL;DR

This study uses Bayesian inference to unify the kinematic analysis of M31's inner and outer globular clusters, testing whether metallicity and substructure distinctions reveal distinct rotational components. By evaluating multiple models (one-, two-, and three-component decompositions) and incorporating metallicity uncertainties, the authors identify Model 2.1 as the most probable, suggesting a fast-rotating, low-metallicity inner subpopulation linked to outer substructures, and a slower, higher-metallicity population aligned with the stellar disk. The results support a multi-epoch accretion history for M31, with different GC subgroups marking distinct assembly events, and highlight the need for broader metallicity data in the outer halo to refine the three-dimensional picture. Overall, the work demonstrates how Bayesian model comparison can disentangle complex kinematic substructures in galaxy halos and illuminate their formation histories.

Abstract

As ancient stellar systems, globular clusters (GCs) offer valuable insights into the dynamical histories of large galaxies. Previous studies of GC populations in the inner and outer regions of the Andromeda Galaxy (M31) have revealed intriguing subpopulations with distinct kinematic properties. Here, we build upon earlier studies by employing Bayesian modelling to investigate the kinematics of the combined inner and outer GC populations of M31. Given the heterogeneous nature of the data, we examine subpopulations defined by GCs' metallicity and by associations with substructure, in order to characterise possible relationships between the inner and outer GC populations. We find that lower-metallicity GCs and those linked to substructures exhibit a common, more rapid rotation, whose alignment is distinct from that of higher-metallicity and non-substructure GCs. Furthermore, the higher-metallicity GCs rotate in alignment with Andromeda's stellar disk. These pronounced kinematic differences reinforce the idea that different subgroups of GCs were accreted to M31 at distinct epochs, shedding light on the complex assembly history of the galaxy.

Figures (10)

• Figure 1: Schematic representation of the metallicity distribution of M31’s inner globular cluster population ($R_p < 25$ kpc, shown in blue) and of the Dulais Structure subset (shown in red), identified by 2023MNRAS.518.5778L using the lowest metallicity inner GCs. While the populations would actually overlap as illustrated here, our modelling uses the simplifying assumption that the overall GC population is neatly split into two at a critical value of metallicity.
• Figure 2: Spatial distribution of M31 GCs, studied in this paper, separated into inner ($R_p < 25$ kpc; left panel) and outer ($R_p \ge 25$ kpc; right panel) populations. Points are colored according to line-of-sight velocity relative to the galaxy, with red indicating positive velocities and blue indicating negative velocities. The size of each point is scaled by the absolute value of the velocity. Black ellipses indicate the disk orientation of the M31 galaxy. Compass arrows indicate the north (N) and east (E) directions.
• Figure 3: The M31 GC population, colour-coded by radial velocity in the M31 frame. The absolute value of the line-of-sight velocity is also indicated by the size of each point. The orange lines are representative samples of the orientation angle drawn from the parameter exploration of Model 1. The ellipse represents M31's disk orientation.
• Figure 4: Corner plot corner of the posterior distribution for Model 1's parameters. There are no strong correlations or other dependencies in the posterior distribution, and all of the marginal distributions are normal to a good approximation. The red solid line represents the median (50th percentile), whereas the blue dashed lines indicate the 16th and 84th percentiles of the posterior distribution of the parameters.
• Figure 5: The posterior distribution for parameters of Model 2.1. Here, $\sigma_1$, $A_1$ and $\phi_1$ are the components for GC$_{\rm non}$ and higher metallicity GCs, while $\sigma_2$, $A_2$ and $\phi_2$ are the components for GC$_{\rm sub}$ and lower metallicity GCs.