paper

Hubble flow variance and the cosmic rest frame

arXiv:1201.5371 · doi:10.1103/PhysRevD.88.083529

Abstract

We characterize the radial and angular variance of the Hubble flow in the COMPOSITE sample of 4534 galaxies, on scales in which much of the flow is in the nonlinear regime. With no cosmological assumptions other than the existence of a suitably averaged linear Hubble law, we find with decisive Bayesian evidence (ln B >> 5) that the Hubble constant averaged in independent spherical radial shells is closer to its asymptotic value when referred to the rest frame of the Local Group, rather than the standard rest frame of the Cosmic Microwave Background. An exception occurs for radial shells in the range 40/h-60/h Mpc. Angular averages reveal a dipole structure in the Hubble flow, whose amplitude changes markedly over the range 32/h-62/h Mpc. Whereas the LG frame dipole is initially constant and then decreases significantly, the CMB frame dipole initially decreases but then increases. The map of angular Hubble flow variation in the LG rest frame is found to coincide with that of the residual CMB temperature dipole, with correlation coefficient -0.92. These results are difficult to reconcile with the standard kinematic interpretation of the motion of the Local Group in response to the clustering dipole, but are consistent with a foreground non-kinematic anisotropy in the distance-redshift relation of 0.5% on scales up to 65/h Mpc. Effectively, the differential expansion of space produced by nearby nonlinear structures of local voids and denser walls and filaments cannot be reduced to a local boost. This hypothesis suggests a reinterpretation of bulk flows, which may potentially impact on calibration of supernovae distances, anomalies associated with large angles in the CMB anisotropy spectrum, and the dark flow inferred from the kinematic Sunyaev-Zel'dovich effect. It is consistent with recent studies that find evidence for a non-kinematic dipole in the distribution of distant radio sources.

37 pages, 9 tables, 13 figures; v2 adds extensive new analysis (including additional subsections, tables, figures); v3 adds a Monte Carlo analysis (with additional table, figure) which further tightens the statistical robustness of the dipole results; v4 adds further clarifications, small corrections, references and discussion of Planck satellite results; v5 typos fixed, matches published version

References in corpus (39)

Cited by in corpus (54)