The global structure of the Milky Way's stellar halo based on the orbits of local metal-poor stars

Abstract: We analyze the global structure of the Milky Way (MW)'s stellar halo including its dominant subcomponent, Gaia-Sausage-Enceladus (GSE). The method to reconstruct the global distribution of this old stellar component is to employ the superposition of the orbits covering over the large MW's space, where each of the orbit-weighting factor is assigned following the probability that the star is located at its currently observed position. The selected local, metal-poor sample with ${\rm [Fe/H]}<-1$ using {\it Gaia} EDR3 and SDSS DR16 shows that the global shape of the halo is systematically rounder at all radii in more metal-poor ranges, such that an axial ratio, $q$, is nearly 1 for ${\rm [Fe/H]}<-2.2$ and $\sim 0.7$ for $-1.4<{\rm [Fe/H]}<-1.0$. It is also found that a halo in relatively metal-rich range of ${\rm [Fe/H]}>-1.8$ actually shows a boxy/peanut-like shape, suggesting a major merger event. The distribution of azimuthal velocities shows a disk-like flattened structure at $-1.4<{\rm [Fe/H]}<-1.0$, which is thought to be the metal-weak thick disk. For the subsample of stars showing GSE-like kinematics and at ${\rm [Fe/H]}>-1.8$, its global density distribution is more spherical with $q \sim 0.9$ than the general halo sample, having an outer ridge at $r\sim20$~kpc. This spherical shape is consistent with the feature of accreted halo components and the ridge suggests that the orbit of GSE's progenitor has an apocenter of $\sim 20$~kpc. Implications for the formation of the stellar halo are also presented.