Breaking the degeneracies between cosmology and galaxy bias

Adopting the framework of the halo occupation distribution ( HOD), we investigate the ability of galaxy clustering measurements to simultaneously constrain cosmological parameters and galaxy bias. Starting with a fiducial cosmological model and galaxy HOD, we calculate spatial clustering observables on a range of length and mass scales, dynamical clustering observables that depend on galaxy peculiar velocities, and the galaxy-matter cross-correlation measurable by weak lensing. We then change one or more cosmological parameters and use chi(2) minimization to find the galaxy HOD that best reproduces the original clustering. Our parameterization of the HOD incorporates a flexible relation between galaxy occupation numbers and halo mass and allows spatial and velocity bias of galaxies within dark matter halos. Despite this flexibility, we find that changes to the HOD cannot mask substantial changes to the matter density Omega(m), the matter clustering amplitude sigma(8), or the shape parameter Gamma of the linear matter power spectrum: cosmology and bias are not degenerate. With the conservative assumption of 10% fractional errors, the set of observables considered here can provide similar to 10% (1 sigma) constraints on sigma 8, Omega(m), and Gamma, using galaxy clustering data alone. The combination sigma(8)Omega(0.75)(m) m is constrained to similar to 5%. In combination with traditional methods that focus on large-scale structure in the "perturbative'' regime, HOD modeling can greatly amplify the cosmological power of galaxy redshift surveys by taking advantage of high-precision clustering measurements at small and intermediate scales ( from sub-Mpc to similar to 20 h(-1) Mpc). At the same time, the inferred constraints on the galaxy HOD provide valuable tests of galaxy formation theory.