The Cores of Dark Matter–Dominated Galaxies: Theory versus Observations

Abstract

We use the rotation curves of a sample of dark matter-dominated dwarf and low surface brightness (LSB) late-type galaxies to study their radial mass distributions. We find that the shapes of the rotation curves are remarkably similar for all (both dwarf and LSB) galaxies in the sample, suggesting a self-similar density distribution of their dark matter (DM) halos. This shape can be reproduced well by a density profile with a shallow central cusp [ρ( r ) ∝ 1/ r γ , γ ≈ 0.2-0.4] corresponding to a steeply rising velocity curve [ v ( r ) ∝ r g , g ≈ 0.9-0.8]. We further show that the observed shapes of the rotation curves are well matched by the average density profiles of dark matter halos formed in very high resolution simulations of the standard cold dark matter model (CDM), the low-density CDM model with cosmological constant (ΛCDM), and the cold + hot dark matter model with two types of neutrino (CHDM). This is surprising in light of several previous studies, which suggested that the structure of simulated dark matter halos is inconsistent with the dynamics of dwarf galaxies. We discuss possible explanations for this discrepancy and show that it is most likely caused by the systematic differences at small radii between the analytic model proposed by Navarro, Frenk, & White, with γ NFW = 1, and the actual central density profiles of the dark matter halos. We also show that the mass distributions in the hierarchically formed halos are on average consistent with the shape of rotation curves of dark matter-dominated galaxies. However, the scatter of the individual profiles around the average is substantial. Finally, we show that the dark matter halos in our hierarchical simulations and the real galaxies in our sample exhibit a similar decrease in their characteristic densities with increasing characteristic radial scales and show increase in their maximum rotation velocities with increase in the radii at which their maximum velocities occur.