2017ApJ...850..116L

We show that the mass of a dark matter halo can be inferred from the dynamical status of its satellite galaxies. Using nine dark matter simulations of halos like the Milky Way (MW), we find that the present-day substructures in each halo follow a characteristic distribution in the phase space of orbital binding energy and angular momentum, and that this distribution is similar from halo to halo, but has an intrinsic dependence on the halo formation history. We construct this distribution directly from the simulations for a specific halo and extend the result to halos of similar formation history but different masses by scaling. The mass of an observed halo can then be estimated by maximizing the likelihood in comparing the measured kinematic parameters of its satellite galaxies with these distributions. We test the validity and accuracy of this method with mock samples taken from the simulations. Using the positions, radial velocities, and proper motions of nine tracers and assuming observational uncertainties comparable to those of MW satellite galaxies, we find that the halo mass can be recovered to within ∼40 % . The accuracy can be improved to within ∼25% if 30 tracers are used. However, the dependence of the phase-space distribution on the halo formation history sets a minimum uncertainty of ∼20 % that cannot be reduced by using more tracers. We believe that this minimum uncertainty also applies to any mass determination for a halo when the phase-space information of other kinematic tracers is used.