2009MNRAS.394L.102L

We model the large kinematic data sets for the four Milky Way dwarf spheroidal (dSph) satellites: Carina, Fornax, Sculptor and Sextans, recently published by Walker et al. The member stars are selected using a reliable dynamical interloper removal scheme tested on simulated data. Our member selection is more restrictive than the one based on metallicity indicators as it removes not only contamination due to Milky Way stars but also the unbound stars from the tidal tails. We model the cleaned data sets by adjusting the solutions of the Jeans equations to the profiles of the projected velocity dispersion and kurtosis. The data are well reproduced by models where mass follows light and the best-fitting stellar orbits are isotropic to weakly tangential, as expected from the tidal stirring scenario. The Fornax dwarf, with more than 2400 member stars, is a dSph galaxy with the most accurately determined mass to date: its 1σ error following from the sampling errors of the velocity moments is below 5 per cent. With mass-to-light ratio of 97 solar units, Sextans seems to be the most dark matter dominated of the four dSph galaxies.