Breaking the mass / anisotropy degeneracy in the Coma cluster

We provide the first direct lifting of the mass/anisotropy degeneracy for a cluster of galaxies, by jointly fitting the line of sight velocity dispersion and kurtosis profiles of the Coma cluster, assuming an NFW tracer density profile, a generalized-NFW dark matter profile and a constant anisotropy profile. We find that the orbits in Coma must be quasi-isotropic, and find a mass consistent with previous analyses, but a concentration parameter 50% higher than expected in cosmological N-body simulations. We then test the accuracy of our method on realistic non-spherical systems with substructure and streaming motions, by applying it to the ten most massive structures in a cosmological N-body simulation. We find that our method yields fairly accurate results on average (within 20%), although with a wide variation (factor 1.7 at 1σ) for the concentration parameter, with decreased accuracy and efficiency when the projected mean velocity is not constant with radius.