The shapes and alignments of the satellites of the Milky Way and Andromeda

Abstract

We measure the intrinsic shapes and alignments of the dwarf spheroidal (dSph) galaxies of the Local Group. We find the dSphs of the Milky Way are intrinsically flatter (mean intrinsic ellipticity $\mu_E\sim0.6$) than those of M31 ($\mu_E\sim0.5$) and that the classical Milky Way dSphs ($M_V<-8.5\,\mathrm{mag}$) are rounder ($\mu_E\sim0.5$) than the ultrafaints ($\mu_E\sim0.65$) whilst in M31 the shapes of the classical and ultrafaint dSphs are very similar. The M31 dSphs are preferentially radially aligned with a dispersion of $\sim45\deg$. This signal is driven by the ultrafaint population whilst the classical M31 dSphs are consistent with a random orientation. We compare our results to the Aquarius mock stellar catalogues of Lowing et al. and find the subhalo radial alignment distribution matches the Local Group dSphs results, whilst the Aquarius intrinsic ellipticities are significantly smaller than the data ($\Delta\langle E\rangle\approx0.4$). We provide evidence that the major axes of the Milky Way satellites lie within a preferential plane with normal vector pointing towards $(\ell,b)=(127,5)\deg$. We associate this preferred direction with the Vast Polar Orbital structure although their respective great circles are offset by $\sim30\deg$. No signal in the alignments of the major axes is found in M31, suggesting that the Great Plane of Satellites is formed from recent accretion or chance alignment. Finally, we provide predictions for the discrepancy between the velocity dispersion versus scale radius distributions for the Milky Way and M31 populations and demonstrate that the projection effect from viewing similar populations from two different locations does not account for the discrepancy which is probably caused by increased tidal disruption in M31.