Globular Cluster Systems

Abstract

1. The basic features of the globular cluster system of the Milky Way are summarized: the total population, subdivision of the clusters into the classic metal-poor and metal-rich components, and first ideas on formation models. The distance to the Galactic center is derived from the spatial distribution of the inner bulge clusters, giving R 0 = (8.1 ± 1.0) kpc. 2. The calibration of the fundamental distance scale for globular clusters is reviewed. Different ways to estimate the zero point and metallicity dependence of the RR Lyrae stars include statistical parallax and Baade-Wesselink measurements of field RR Lyraes, astrometric parallaxes, white dwarf cluster sequences, and field subdwarfs and main sequence fitting. The results are compared with other distance measurements to the Large Magellanic Cloud and M31. 3. Radial velocities of the Milky Way clusters are used to derive the kinematics of various subsamples of the clusters (the mean rotation speed about the Galactic center, and the line-of-sight velocity dispersion). The inner metal-rich clusters behave kinematically and spatially like a flattened rotating bulge population, while the outer metal-rich clusters resemble a thick-disk population more closely. The metal-poor clusters have a significant prograde rotation (80 to 100 km s−1) in the inner halo and bulge, declining smoothly to near-zero for R ≳ R 0. No identifiable subgroups are found with significant retrograde motion. 4. The radial velocities of the globular clusters are used along with the spherically symmetric collisionless Boltzmann equation to derive the mass profile of the Milky Way halo. The total mass of the Galaxy is near ≃ 8 × 1011 M⊙ for r ≲ 100 kpc. Extensions to still larger radii with the same formalism are extremely uncertain because of the small numbers of outermost satellites, and the possible correlations of their motions in orbital families. 5. The luminosity functions (GCLF) of the Milky Way and M31 globular clusters are defined and analyzed. We search for possible trends with cluster metallicity or radius, and investigate different analytic fitting functions such as the Gaussian and power-law forms. 6. The global properties of GCSs in other galaxies are reviewed. Measureable distributions include the total cluster population (quantified as the specific frequency S N), the metallicity distribution function (MDF), the luminosity and space distributions, and the radial velocity distribution. 7. The GCLF is evaluated as a standard candle for distance determination. For giant E galaxies, the GCLF turnover has a mean luminosity of M v = −7.33 on a distance scale where Virgo has a distance modulus of 31.0 and Fornax is at 31.3, with galaxy-to-galaxy scatter σ(M v) = 0.15 mag. Applying this calibration to more remote galaxies yields a Hubble constant H 0 = (74 ± 9) km s−1 Mpc−1. 8. The observational constraints on globular cluster formation models are summarized. The appropriate host environments for the formation of ∼ 105 to 106 M⊙ clusters are suggested to be kiloparsec-sized gas clouds (Searle/Zinn fragments or supergiant molecular clouds) of 108 to 109 M⊙. A model for the growth of proto-cluster clouds by collisional agglomeration is presented, and matched with observed mass distribution functions. The issue of globular cluster formation efficiency in different galaxies is discussed (the “specific frequency problem”). 9. Other influences on galaxy formation are discussed, including mergers, accretions, and starbursts. Mergers of disk galaxies almost certainly produce elliptical galaxies of low S n, while the high-S N ellipticals are more likely to have been produced through in situ formation. Starburst dwarfs and large active galaxies in which current globular cluster formation is taking place are compared with the key elements of the formation model. 10. (Appendix) Some basic principles of photometric methods are gathered together and summarized: the fundamental signal-to-noise formula, objective star finding, aperture photometry, PSF fitting, artificial-star testing, detection completeness, and photometric uncertainty. Lastly, we raise the essential issues in photometry of nonstellar objects, including image moment analysis, total magnitudes, and object classification techniques.