Physical bias of galaxies from large-scale hydrodynamic simulations

We analyze a new large-scale (100 h(-1) Mpc) numerical hydrodynamic simulation of the popular ACDM cosmological model, including in our treatment dark matter, gas, and star formation, on the basis of standard physical processes. The method, applied with a numerical resolution of <200 h(-1) kpc (which is still quite coarse for following individual galaxies, especially in dense regions), attempts to estimate where and when galaxies form. We then compare the smoothed galaxy distribution with the smoothed mass distribution to determine the "bias," defined as b = (delta M/M)(gal)/(delta M/M)(total), on scales that are large compared to the code numerical resolution ton the basis of resolution tests given in the Appendix of this paper). We find that (holding all variables constant except the quoted one) bias increases with decreasing scale, with increasing galactic age or metallicity, and with increasing redshift of observations. At the 8 h(-1) Mpc fiducial comoving scale, bias (for bright regions) is 1.35 at z = 0, reaching to 3.6 at z = 3, both numbers being consistent with extant observations. We also find that (10-20) h(-1) Mpc voids in the distribution of luminous objects are as observed (i.e., observed voids are not an argument against cold dark matter [CDM]-like models), and finally that the younger systems should show a colder Hubble how than do the early-type galaxies (a testable proposition). Surprisingly, little evolution is found in the amplitude of the smoothed galaxy-galaxy correlation function las a function of comoving separation). Testing this prediction against observations will allow a comparison between this work and that of Kauffmann et al., which is based on a different physical modeling method.