The asymmetric distribution of satellite galaxy velocities

Companion galaxies tend to have redshifts that exceed that of their associated primary. We show that the latest sample of satellite galaxies (from Zaritsky et al.) corroborates this assertion. We examine whether selection biases and contamination by galaxies that are not physical satellites can create the observed velocity asymmetry. As shown by Valtonen and Byrd for previous samples, simple models of the selection process reproduce the sense of the asymmetry, but not its magnitude unless the samples are dominated by interlopers. We expand on this model by using numerical simulations that include a simple model of galaxy clustering, and demonstrate that clustering can increase the magnitude of the expected asymmetry for a particular level of contamination. Depending on input assumptions, these models imply 1 a lower limits on the contamination level in the Zaritsky et al. sample of between 8% and 20%. For an independent estimate of the characteristics of interlopers, we obtained a pure sample of interlopers from the CfA Redshift Catalogue. Using the results from the analysis of that sample and the models, we conclude that, for the type of satellite galaxy sample discussed here, a moderate level of contamination (∼ 10%) is sufficient to account for the observed asymmetry.