# Evidence for an environment-dependent shift in the baryon acoustic oscillation peak Article - Avril 2015

Boudewijn F. Roukema, Thomas Buchert, Jan J. Ostrowski, Martin J. France

Boudewijn F. Roukema, Thomas Buchert, Jan J. Ostrowski, Martin J. France, « Evidence for an environment-dependent shift in the baryon acoustic oscillation peak  », Monthly Notices of the Royal Astronomical Society, avril 2015, pp. 1660 - 1673. ISSN 0035-8711

Abstract

The Friedmann-Lemaitre-Robertson-Walker (FLRW) metric assumes comoving spatial rigidity of metrical properties. The curvature term in comoving coordinates is environment-independent and cannot evolve. In the standard model, structure formation is interpreted accordingly : structures average out on the chosen metrical background, which remains rigid in comoving coordinates despite nonlinear structure growth. The latter claim needs to be tested, since it is a hypothesis that is not derived using general relativity. We introduce a test of the comoving rigidity assumption by measuring the two-point auto-correlation function on comoving scales---assuming FLRW comoving spatial rigidity---in order to detect shifts in the baryon acoustic oscillation (BAO) peak location for Large Red Galaxy (LRG) pairs of the Sloan Digital Sky Survey Data Release 7. In tangential directions, subsets of pairs overlapping with superclusters or voids show the BAO peak. The tangential BAO peak location for overlap with Nadathur & Hotchkiss superclusters is $4.3\pm1.6$ Mpc/h less than that for LRG pairs unselected for supercluster overlap, and $6.6\pm2.8$ Mpc/h less than that of the complementary pairs. Liivamagi et al. superclusters give corresponding differences of $3.7\pm2.9$ Mpc/h and $6.3\pm2.6$ Mpc/h, respectively. We have found moderately significant evidence (Kolmogorov—Smirnov tests suggest very significant evidence) that the BAO peak location for supercluster-overlapping pairs is compressed by about 6% compared to that of the complementary sample, providing a potential challenge to FLRW models and a benchmark for predictions from models based on an averaging approach that leaves the spatial metric a priori unspecified.