The clustering of galaxies in the SDSS-III DR9 Baryon Oscillation Spectroscopic Survey: testing deviations from Λ and general relativity using anisotropic clustering of galaxies

Abstract

We use the joint measurement of geometry and growth from anisotropic galaxy\nclustering in the Baryon Oscillation Spectroscopic Survey Data Release 9 CMASS\nsample reported by Reid et al. to constrain dark energy properties and possible\ndeviations from the General Relativity. Assuming GR and taking a prior on the\nlinear matter power spectrum at high redshift from the cosmic microwave\nbackground (CMB), anisotropic clustering of the CMASS DR9 galaxies alone\nconstrains $\\Omega_{\\rm m} = 0.308 \\pm 0.022$ and $100\\Omega_{\\rm k} = 5.9 \\pm\n4.8$ for $w = -1$, or $w = -0.91 \\pm 0.12$ for $\\Omega_k = 0$. When combined\nwith the full CMB likelihood, the addition of the anisotropic clustering\nmeasurements to the spherically-averaged BAO location increases the\nconstraining power on dark energy by a factor of 4 in a flat CDM cosmology with\nconstant dark energy equation of state $w$ (giving $w = -0.87 \\pm 0.05$). This\nimpressive gain depends on our measurement of both the growth of structure and\nAlcock-Paczynski effect, and is not realised when marginalising over the\namplitude of redshift space distortions. Combining with both the CMB and\nSupernovae Type Ia (SNeIa), we find $\\Omega_{\\rm m} = 0.281 \\pm 0.014$ and\n$1000\\Omega_{\\rm k}=-9.2\\pm5.0$ for $w = -1$, or $w_0 = -1.13 \\pm 0.12$ and\n$w_{\\rm a}=0.65 \\pm 0.36$ assuming $\\Omega_k = 0$. Finally, when a $\\Lambda$CDM\nbackground expansion is assumed, the combination of our estimate of the growth\nrate with previous growth measurements provides tight constraints on the\nparameters describing possible deviations from GR giving $\\gamma = 0.64 \\pm\n0.05$. For one parameter extensions of the flat $\\Lambda$CDM model, we find a\n$\\sim 2\\sigma$ preference either for $w > -1$ or slower growth than in GR.\nHowever, the data is fully consistent with the concordance model, and the\nevidence for these additional parameters is weaker than $2\\sigma$.\n