Klaudia Kowalczyk, Ewa L. Lokas, Stelios Kazantzidis, Lucio Mayer
Using collisionless N-body simulations of dwarf galaxies orbiting the Milky Way (MW) we construct realistic models of dwarf spheroidal (dSph) galaxies of the Local Group. The dwarfs are initially composed of stellar disks embedded in dark matter haloes with different inner density slopes and are placed on an eccentric orbit typical for MW subhaloes. After a few Gyr of evolution the stellar component is triaxial as a result of bar instability induced by tidal forces. Observing the simulated dwarfs along three principal axes of the stellar component we create mock data sets and determine the their half-light radii and line-of-sight velocity dispersions. Using the estimator proposed by Wolf et al. we calculate masses within half-light radii. The masses obtained this way are over(under)estimated by up to a factor of two when the line of sight is along the longest (shortest) axis of the stellar component. We then divide the initial stellar distribution into an inner and outer population and trace their evolution in time. The two populations, although affected by tidal forces, retain different density profiles even after a few Gyr. We measure the half-light radii and velocity dispersions of the stars in the two populations along different lines of sight and use them to estimate the slope of the mass distribution in the dwarfs following the method proposed by Walker & Penarrubia. The inferred slopes are systematically over- or underestimated, depending on the line of sight. In particular, when the dwarf is seen along the longest axis of the stellar component, a significantly shallower density profile is inferred than the real one measured from the simulations. Since most dSphs are non-spherical and their orientation with respect to our line of sight is unknown, the method can be reliably applied only to a large sample of dwarfs when these systematic errors are expected to be diminished.
View original:
http://arxiv.org/abs/1212.3438
No comments:
Post a Comment