5.1. Dwarf Spheroidal Galaxies and Dark Matter

Galactic dSphs are of particular interest in efforts to elucidate the nature of dark matter since they may be dark-matter-dominated and can be studied in great detail due to their proximity. From an analysis of the kinematic properties of Draco and Ursa Minor Gerhard & Spergel (1992a) exclude fermionic light particles (neutrinos) as dark matter suspects because phase-space limits would then require unreasonably large core radii and masses for these two galaxies.

The initial measurements of velocity dispersions in dSphs were criticized for including luminous AGB stars and Carbon stars, whose radial velocities may reflect atmospheric motions, and for neglecting the impact of binaries (see Olszewski 1998 for details). Subsequent studies concentrated on somewhat fainter stars along the upper RGB, carried out extensive simulations to assess the impact of binaries (Hargreaves, Gilmore, & Annan 1996; Olszewski, Pryor, & Armandroff 1996), obtained multi-epoch observations (e.g., Olszewski, Aaronson, & Hill 1995), and increased the number of red giants with measured radial velocities to more than 90 in some cases (Armandroff, Olszewski, & Pryor 1995). These studies established that the large velocity dispersions in dSphs are not due to the previously mentioned observational biases. Kleyna et al. (1999) show that the currently available measurements for the two best-studied dSphs, Draco and Ursa Minor, are not yet sufficient to distinguish between models where mass follows light (constant M/L throughout the dSph) or extended dark halo models when interpreting the velocity dispersions as high M/L ratios due to large dark matter content. Mateo (1998) and Mateo et al. (1998) argue that the relation between total M/L and V-band luminosity for dSphs can be approximated well when adopting a stellar M/L of 1.5 (similar to globular clusters) and an extended dark halo with a mass of 2 . 10⁷ M, suggesting fairly uniform properties for the dark halos of dSphs.

Luminosity functions (LFs) of old stellar systems can provide further constraints on the nature of dark matter. The main-sequence LFs of old field populations in the Galactic bulge (Holtzman et al. 1998), LMC and SMC (Holtzman et al. 1999), Draco (Grillmair et al. 1998), and Ursa Minor (Feltzing et al. 1999) are in excellent agreement with the solar neighborhood IMF and LFs of globular clusters that did not suffer mass segregation. Since globular clusters are not known to contain dark matter, one would expect to find differences in the LF of dark-matter-rich populations if low-mass objects down to 0.45 M were important contributors to the baryonic dark matter content. Furthermore, these studies demonstrate that the LF in objects with a wide range of M/L ratios does not differ much. The possible contribution of white dwarfs (or lack thereof) is discussed elsewhere in these proceedings.