On the Convergence of the Milky Way and M31 Kinematics from Cosmological Simulations

TL;DR

This paper investigates how the size of cosmological simulations affects the accuracy of modeling the Local Group's kinematics, emphasizing the need for large simulation boxes to achieve reliable results.

Contribution

It demonstrates that large simulation volumes (~1 Gpc) are necessary for convergence in Local Group kinematic properties, providing updated velocity estimates for M31 relative to the Milky Way.

Findings

LG velocities are consistent with recent proper motion measurements.

Simulation box size critically impacts the derived kinematic properties.

Large simulations reduce biases in Local Group analogues.

Abstract

The kinematics of the Milky Way (MW) and M31, the dominant galaxies in the Local Group (LG), can be used to estimate the LG total mass. New results on the M31 proper motion have recently been used to improve that estimate. Those results are based on kinematic priors that are sometimes guided and evaluated using cosmological N-body simulations. However, the kinematic properties of simulated LG analogues could be biased due to the effective power spectrum truncation induced by the small size of the parent simulation. Here we explore the dependence of LG kinematics on the simulation box size to argue that cosmological simulations need a box size on the order of 1 Gpc in order to claim convergence on the LG kinematic properties. Using a large enough simulation, we find M31 tangential and radial velocities relative to the MW to be in the range $v_{\mathrm {tan}}=105^{+94}_{-59} $ km/s and $v_{\mathrm {rad}}=-108^{+68}_{-81}$ km/s, respectively. This study highlights that LG kinematics derived from N-body simulations have to be carefully interpreted taking into account the size of the parent simulation.