Dark galaxies, spin bias and gravitational lenses

Abstract

Gravitational lensing studies suggest that the Universe may contain a population of dark galaxies; we investigate this intriguing possibility and propose a mechanism to explain their nature. In this mechanism a dark galaxy is formed with a low density disk in a dark halo of high spin parameter; such galaxies can have surface densities below the critical Toomre value for instabilities to develop, and following Kennicutt's work we expect these galaxies to have low star formation rates. The only stellar component of the galaxies is a halo system, formed during the collapse of the proto-galactic cloud. We compute synthetic stellar population models and show that, at a redshift $z=0.5$, such galaxies have apparent magnitudes $B \simeq 28, R \simeq 26$ and $I \simeq 25$, and could be unveiled by deep searches with the Hubble Space Telescope. Dark galaxies have an initial short blue phase and then become essentially invisible, therefore they may account for the blue population of galaxies at high redshift. We find a strong mass-dependence in the fraction of dark galaxies, and predict that spiral galaxies will not be found in halos with masses less than about $10^9 M_\odot$, if $\Omega=1$. Above about $10^{12}M_\odot$, all halos can produce luminous disks. The mass-dependence of the galaxy-formation efficiency introduces the possibility of `spin bias' -- luminous galaxies being associated preferentially with strongly-clustered high-mass halos. A further prediction is that the slope of the faint-end luminosity function for galaxies will be flatter than the associated halo mass function.