Testing Gravity Using Large-Scale Redshift-Space Distortions

TL;DR

This paper tests alternative cosmological models to LCDM+GR using large-scale redshift-space distortions from SDSS data, constraining model parameters through clustering measurements and likelihood analysis.

Contribution

It introduces a parameterized growth factor formula for Unified Dark Matter models and applies a refined methodology to constrain alternative gravity models using SDSS clustering data.

Findings

UDM models have a speed of sound less than 6.1e-4 at 95% confidence.

nDGP model's crossover scale r_c exceeds 340 Mpc at 95% confidence.

Accurate growth rate measurements must account for wide-angle effects.

Abstract

We use Luminous Red Galaxies from the Sloan Digital Sky Survey II to test the cosmological structure growth in two alternatives to the standard LCDM+GR cosmological model. We compare observed three-dimensional clustering in SDSS DR7 with theoretical predictions for the standard vanilla LCDM+GR model, Unified Dark Matter cosmologies and the normal branch DGP. In computing the expected correlations in UDM cosmologies, we derive a parameterized formula for the growth factor in these models. For our analysis we apply the methodology tested in Raccanelli et al. 2010 and use the measurements of Samushia et al. 2011, that account for survey geometry, non-linear and wide-angle effects and the distribution of pair orientation. We show that the estimate of the growth rate is potentially degenerate with wide-angle effects, meaning that extremely accurate measurements of the growth rate on large scales will need to take such effects into account. We use measurements of the zeroth and second order moments of the correlation function from SDSS DR7 data and the Large Suite of Dark Matter Simulations, and perform a likelihood analysis to constrain the parameters of the models. Using information on the clustering up to r_max = 120 Mpc/h, and after marginalizing over the bias, we find, for UDM models, a speed of sound < 6.1e-4, and, for the nDGP model, a cross-over scale r_c > 340 Mpc, at 95% confidence level.