Institution(s): 1. University of Oxford
Halo stars may have formed in situ, been flung out from further in, or accreted from neighbouring systems. The particular combination of processes is embedded in the phase-space and metallicity distributions of the stars, and therefore they serve as fossils of the halo's formation history. Mutual investigation of these distributions is paramount to make the most of the observations available to us.
We examine the existence of a double power-law density profile found in several recent studies of stellar populations from a full chemodynamical perpective by extending a recently proposed action-based distribution function (Posti et al., 2015, Williams et al., 2015) to also vary continuously as a function of metallicity. We infer the parameters of the extended distribution function by maximising the likelihood of phase-space and metallicity observations of a sample of K giants, blue horizontal branch stars, and RR Lyrae that trace the halo out to about 120kpc. We assume the stars to be moving in a potential consisting of thin, thick, and gas disc, bulge, and dark matter halo components. A distribution function is recovered that is able to satisfactorily fit the data, implying that there is a smooth component in the halo exhibiting equilibrium dynamics. The best-fit parameters describe a double-power law density profile similar to that found in past studies. In particular, we learn that the break is supported dynamically by apocentres preferentially being located there (a mechanism proposed to explain the Milky Way's break radius from accretion only). The halo is found to be radially anisotropic, and more so in the metal-poorer stars, supporting the contribution of accretion events to the build-up of the halo.