2024MNRAS.533.3426M

We analyse the spatial anisotropy and the velocity anisotropy in a set of mock stellar haloes from the Aquarius simulations. The spatial anisotropy in each mock stellar halo rises progressively with the increasing distance from the halo centre, eventually reaching a maximum near the periphery. Excluding the bound satellites leads to a significant reduction of the spatial anisotropy in each halo. We compare the measured anisotropy in the mock stellar haloes with that from their sphericalized versions where all the shape- and substructure-induced anisotropies are erased. The growth of spatial anisotropy persists throughout the entire halo when the bound satellites are present but remains limited within the inner halo (${\lt} 60 \, {h^{-1}\, {\rm kpc}}$) after their exclusion. This indicates that the spatial anisotropy in the inner halo is induced by the diffuse substructures and the halo shape, whereas the outer halo anisotropy is dominated by the bound satellites. We find that the outer parts of the stellar halo are kinematically colder than the inner regions. The stellar orbits are predominantly radial, but they become rotationally dominated at certain radii that are marked by the prominent dips in the velocity anisotropy. Most of these dips disappear after the removal of the satellites. A few shallow dips arise occasionally due to the presence of diffuse streams and clouds. Our analysis suggests that a combined study of the spatial and velocity anisotropies can reveal the structure and the assembly history of the stellar haloes.