We use the high‐resolution cosmological N‐body simulations from the Aquarius project to investigate in detail the mechanisms that determine the shape of Milky Way type dark matter haloes. We find that, when measured at the instantaneous virial radius, the shape of individual haloes changes with time, evolving from a typically prolate configuration at early stages to a more triaxial/oblate geometry at the present day. This evolution in halo shape correlates well with the distribution of the infalling material: prolate configurations arise when haloes are fed through narrow filaments, which characterizes the early epochs of halo assembly, whereas triaxial/oblate configurations result as the accretion turns more isotropic at later times. Interestingly, at redshift z= 0, clear imprints of the past history of each halo are recorded in their shapes at different radii, which also exhibit a variation from prolate in the inner regions to triaxial/oblate in the outskirts. Provided that the Aquarius haloes are fair representatives of Milky Way like 1012 M⊙ objects, we conclude that the shape of such dark matter haloes is a complex, time‐dependent property, with each radial shell retaining memory of the conditions at the time of collapse.

We study the radial migration of stars driven by recurring multi-arm spiral features in an exponential disk embedded in a dark matter halo. The spiral perturbations redistribute angular momentum within the disk and lead to substantial radial displacements of individual stars, in a manner that largely preserves the circularity of their orbits and that results, after 5 Gyr (∼ 40 full rotations at the disk scalelength), in little radial heating and no appreciable changes to the vertical or radial structure of the disk. Our results clarify a number of issues related to the spatial distribution and kinematics of migrators. In particular, we find that migrators are a heavily biased subset of stars with preferentially low vertical velocity dispersions. This "provenance bias" for migrators is not surprising in hindsight, for stars with small vertical excursions spend more time near the disk plane and thus respond more readily to non-axisymmetric perturbations. We also find that the vertical velocity dispersion of outward migrators always decreases, whereas the opposite holds for inward migrators. To first order, newly arrived migrators simply replace stars that have migrated off to other radii, thus inheriting the vertical bias of the latter. Extreme migrators might therefore be recognized, if present, by the unexpectedly small amplitude of their vertical excursions. Our results show that migration, understood as changes in angular momentum that preserve circularity, can affect strongly the thin disk, but cast doubts on models that envision the Galactic thick disk as a relic of radial migration.

We propose a new model for the dark matter halo of the Milky Way that fits the properties of the stellar stream associated with the Sagittarius dwarf galaxy. Our dark halo is oblate with q z = 0.9 for r 10 kpc, and can be made to follow the Law & Majewski model at larger radii. However, we find that the dynamical perturbations induced by the Large Magellanic Cloud on the orbit of Sgr cannot be neglected when modeling its streams. When taken into account, this leads us to constrain the Galaxy's outer halo shape to have minor-to-major axis ratio (c/a) Φ = 0.8 and intermediate-to-major axis ratio (b/a) Φ = 0.9, in good agreement with cosmological expectations.

We present a new analysis of the Aquarius simulations done in combination with a semi-analytic galaxy formation model. Our goal is to establish whether the subhalos present in ΛCDM simulations of Milky Way-like systems could host the dwarf spheroidal (dSph) satellites of our Galaxy. Our analysis shows that, contrary to what has been assumed in most previous work, the mass profiles of subhalos are generally not well fit by NFW models but that Einasto profiles are preferred. We find that for shape parameters α = 0.2 − 0.5 and v max = 10 − 30 km/s there is very good correspondence with the observational constraints obtained for the nine brightest dSph of the Milky Way. However, to explain the internal dynamics of these systems as well as the number of objects of a given circular velocity the total mass of the Milky Way should be ∼ 8 × 10 11 M , a value that is in agreement with many recent determinations, and at the low mass end of the range explored by the Aquarius simulations. Our simulations show important scatter in the number of bright satellites, even when the Aquarius Milky Way-like hosts are scaled to a common mass, and we find no evidence for a missing population of massive subhalos in the Galaxy. This conclusion is also supported when we examine the dynamics of the satellites of M31.