Published Article: The Local Group’s mass: probably no more than the sum of its parts
Our article describing how the mass of the Local Group and the mass of its two most massive members (the Milky Way and M31) can be constrained using the extremely large UCHUU simulation and machine learning.
Abstract: The total mass of the Local Group and the masses of its primary constituents, the Milky Way (MW) and M31, are important anchors for several cosmological questions. Recent independent measurements have consistently yielded halo masses close to 1012 M☉ for the MW, and 1-2×1012 M☉ for M31, while estimates derived from the pair’s kinematics via the ‘timing argument’ have yielded a combined mass of around 5×1012 M☉ . We analyse the extremely large UCHUU simulation to constrain the mass of the Local Group and its two most massive members. First, we demonstrate the importance of selecting pairs whose kinematics reflect their mutual interactions. Adopting the observed separation and radial velocity, we obtain a weighted posterior of 75 (+65-40) km s-1 for the uncertain transverse velocity. Via Gaussian process regression, we infer a total mass of 3.2 (+1.2-0.9)×1012 M☉, significantly below the timing argument value. Importantly, the remaining uncertainty is not rooted in the analysis or observational errors, but in the irreducible scatter in the kinematics-mass relation. We further find a mass for the less massive halo of 0.9(+0.6-0.3)×1012 M☉ and for the more massive halo of 2.3(+1.0-0.9)×1012 M☉ , consistent with independent measurements of the masses of MW and M31, respectively. Incorporating the MW mass as an additional prior let us constrain all measurements further and determine that the MW is very likely less massive than M31.