Cosmological velocity correlations - Observations and model predictions

Abstract

view Abstract Citations (126) References (42) Co-Reads Similar Papers Volume Content Graphics Metrics Export Citation NASA/ADS Cosmological Velocity Correlations: Observations and Model Predictions Gorski, Krzysztof M. ; Davis, Marc ; Strauss, Michael A. ; White, Simon D. M. ; Yahil, Amos Abstract We introduce simple statistics for measuring two-point cosmological peculiar velocity correlations, using the observed radial velocities of galaxies and estimates of their distances. We apply these statistics to two real data sets: the Local Supercluster spiral galaxy sample of Aaronson et al. and the elliptical galaxy sample of Burstein et al., and to the velocity field predicted by the distribution of IRAS galaxies. The coherence length, defined as the separation at which the correlations drop to half their zero lag value, is between 1100 and 1600 km s^-1^ in these three samples. The derived correlations for the IRAS velocity field are very insensitive to the distribution of test points probing the field. The velocity correlations of the Local Supercluster galaxy sample appear consistent with those of the IRAS velocity field, but the elliptical galaxy sample gives systematically higher velocity correlations at small separations. However, this amplitude is highly dependent on the peculiar velocities assigned to the 11 galaxies in three clusters in the Centaurus direction; if these three clusters are assigned peculiar velocities of 500 km s^-1^, rather than the observed values of over 1000 km s^-1^, the velocity correlations in the elliptical sample become consistent with those derived for the other two samples. We have compared these results with predictions from two models of the formation of large-scale structure: cold dark matter (CDM) and Peebles's baryon isocurvature (PBI). We have performed N-body simulations of these models to check the linear theory predictions and to measure sampling fluctuations. The CDM linear theory and N-body predictions for the coherence length agree rather well with the measured values of the data samples. However, the amplitudes of the observed velocity correlations are better matched if the matter spatial correlations are stronger than has been previously assumed. Thus only a few percent of observers in a b = 2.5 CDM universe, where b is the bias parameter, would measure a velocity correlation amplitude as large as that of the Local Supercluster and IRAS samples, and none would observe an amplitude as large as is measured for the (unaltered) elliptical galaxy sample. in a b = 2.0 CDM model, ~30% of observers would measure a correlation amplitude as large as that of the Local Supercluster and IRAS samples, but the elliptical galaxy sample would still be an extremely rare event. The typical PBI realization shows much greater long-range coherence and amplitude than are seen in the data sets. Even so, a few percent of observers in such a universe would measure velocity correlations consistent with the reported elliptical galaxy data. We demonstrate with the N-body models the sensitivity of the velocity correlations to distance errors. In a sample selected from a uniform galaxy distribution, distance errors give rise to a monopole peculiar velocity outflow, which can be readily removed in a statistical fashion. There is, however, an additional systematic effect due to density inhomogeneities, which cannot be removed statistically. It is conceivable that such an effect could be partly responsible for the observed differences in the correlations derived for the data sets. Publication: The Astrophysical Journal Pub Date: September 1989 DOI: 10.1086/167771 Bibcode: 1989ApJ...344....1G Keywords: Astronomical Models; Cosmology; Galactic Clusters; Radial Velocity; Dark Matter; Elliptical Galaxies; Many Body Problem; Spiral Galaxies; Astrophysics; COSMOLOGY; DARK MATTER; GALAXIES: CLUSTERING full text sources ADS | data products SIMBAD (2) NED (1)