Where Do Galaxies End?

Abstract

Our current view of galaxies considers them as systems of stars and gas embedded in extended halos of dark matter, much of it formed by the infall of smaller systems at earlier times. The true extent of a galaxy remains poorly determined, with the virial radius (R_vir) providing a characteristic separation between collapsed structures in dynamical equilibrium and external infalling matter. Other physical estimates of the extent of gravitational influence include the gravitational radius, gas accretion radius, and "galactopause" arising from outflows that stall at 100-200 kpc over a range of outflow parameters and confining gas pressures. Physical criteria are proposed to define bound structures, including a more realistic definition of R_vir (M*, M_h, z_a) for stellar mass M* and halo mass M_h, half of which formed at "assembly redshifts" z_a = 0.7-1.3. We estimate the extent of bound gas and dark matter around L* galaxies to be ~200 kpc. The new virial radii, with mean R_vir ~ 200 kpc, are 40-50% smaller than values estimated in recent HST/COS detections of H I and O VI absorbers around galaxies. In the new formalism, the Milky Way stellar mass, log M* = 10.7 +/- 0.1, would correspond to R_vir = 153 (+25,-16) kpc for half-mass halo assembly at z_a = 1.06 +/- 0.03. The frequency per unit redshift of low-redshift O VI absorption lines in QSO spectra suggests absorber sizes ~150 kpc when related to intervening 0.1L* galaxies. This formalism is intended to clarify semantic differences arising from observations of extended gas in galactic halos, circumgalactic medium (CGM), and filaments of the intergalactic medium (IGM). Astronomers should refer to bound gas in the galactic halo or CGM, and unbound gas at the CGM-IGM interface, on its way into the IGM.