Clustering and bias measurements of SDSS voids

Abstract

Using a void catalogue from the Sloan Digital Sky Survey, we present the first measurements of void clustering and the corresponding void bias. Over the range 30-200 Mpc h-1, the void autocorrelation is detected at 5σ significance for voids of radius 15-20 Mpc h-1. We also measure the void-galaxy cross-correlation at higher signal to noise and compare the inferred void bias with the autocorrelation results. Void bias is constant with scale for voids of a given size, but its value falls from 5.6 ± 1.0 to below zero as the void radius increases from 15 to 30 Mpc h-1. The comparison of our measurements with carefully matched galaxy mock catalogues, with no free parameters related to the voids, shows that model predictions can be reliably made for void correlations. We study the dependence of void bias on tracer density and void size with a view to future applications. In combination with our previous lensing measurements of void mass profiles, these clustering measurements provide another step towards using voids as cosmological tracers.