Simulating the Anisotropic Clustering of Luminous Red Galaxies with Subhalos: A Direct Confrontation with Observation and Cosmological Implications

Abstract: We model the apparent clustering anisotropy of Luminous Red Galaxies (LRGs) in the Sloan Digital Sky Survey using subhalos identified in cosmological $N$-body simulations. We first conduct a Markov-chain Monte Carlo analysis on the parameters characterizing subhalos hosting LRGs assuming a specific $\Lambda$CDM cosmology on which we run the simulations. We show that simple models with central and satellite subhalos can explain the observed multipole moments of the power spectrum up to hexadecapole on large scales ($k\lesssim0.3~h\mathrm{Mpc}^{-1}$). A satellite fraction of $20$ to $30$ per cent is favored weakly depending on the detail of the model. The fraction is shown to be robust when we adopt a more refined model based on the halo occupation number from the literature. We then vary cosmological parameters controlling the anisotropy in redshift-space effectively by deforming the simulation box (the Alcock-Paczynski effect) and changing the amplitude of the velocities (the redshift-space distortions). We demonstrate that we can constrain the geometry of the universe, the structure growth rate, and the parameters characterizing LRGs simultaneously. This is a step toward cosmological analysis with realistic bias description beyond empirical bias functions with nuisance parameters.