2024MNRAS.532.2490S

The mass distribution in the Local Group (LG), dominated by the Andromeda (M31) and Milky Way (MW) pair, is highly anisotropic. We use the APOSTLE simulations to examine how this anisotropy manifests on the spatial distribution and kinematics of dwarf galaxies out to a distance of $d_{\rm MW}\sim 3$ Mpc from the MW. The simulations indicate a preference for dwarfs to be located near the axis defined by the MW-M31 direction, even for dwarfs in the LG periphery (LGP; i.e. at distances $1.25\ \lt\ d_{\rm MW}/$Mpc $\lt\ 3$). The 'Hubble flow' in the periphery is also affected; at fixed $d_{\rm MW}$ the mean recession speed, $\langle V_{\rm rad} \rangle$, varies with angular distance to M31, peaking in the anti-M31 direction and reaching a minimum behind M31. The M31-MW mass decelerates the local expansion; the LG 'turnaround radius' (i.e. where $\langle V_{\rm rad} \rangle =0$) in APOSTLE is at $r \sim 1.25$ Mpc from the LG barycentre and the pure Hubble flow (where $\langle V_{\rm rad} \rangle \sim H_0*d$) is reached beyond $r\sim 3$ Mpc. The predicted flow is very cold, with a barycentric dispersion of $\lt 40$ km s^–1. Comparing these predictions with observations yields mixed results. There is little evidence for a preferred alignment of dwarfs along the MW-M31 direction, but some evidence for an angular anisotropy in $\langle V_{\rm rad} \rangle$. Although the 'coldness' of the Hubble flow is consistent with the simulations, it is less decelerated: relative to the MW all galaxies beyond $d_{\rm MW} \sim 1.25$ Mpc seem to be already on a pure Hubble flow. We argue that these oddities may result at least partly from incompleteness and inhomogeneous sky coverage in our current inventory of nearby dwarfs.