Testing cosmology with a catalogue of voids in the BOSS galaxy surveys

We present a public catalogue of voids in the Baryon Oscillation Spectroscopic Survey (BOSS) Data Release 11 LOWZ and CMASS galaxy surveys. This catalogue contains information on the location, sizes, densities, shapes and bounding surfaces of 8956 independent, disjoint voids, making it the largest public void catalogue to date. Voids are identified using a version of the ZOBOV algorithm, the operation of which has been calibrated through tests on mock galaxy populations in N-body simulations, as well as on a suite of 4096 mock catalogues which fully reproduce the galaxy clustering, survey masks and selection functions. Based on this, we estimate a false positive detection rate of 3%. Comparison with mock catalogues limits deviations of the void size distribution from that predicted in the $\Lambda$CDM model to be less than 6% for voids with effective radius $8<R_v<60\,h^{-1}$Mpc and in the redshift range $0.15<z<0.7$. This could tightly constrain modified gravity scenarios and models with a varying equation of state, but we identify systematic biases which must be accounted for to reduce the theoretical uncertainty in the predictions for these models to the current level of precision attained from the data. We also examine the distribution of void densities and identify a deficit of the deepest voids relative to $\Lambda$CDM expectations, which is significant at more than the $3\sigma$ equivalent level. We discuss possible explanations for this discrepancy but at present its cause remains unknown.