Bianchi I meets the Hubble diagram

Abstract

We improve existing fits of the Bianchi I metric to the Hubble diagram of supernovae and find an intriguing yet non-significant signal for anisotropy that should be verified or falsified in the near future by the Large Synoptic Survey Telescope. Since the literature contains two different formulas for the apparent luminosity as a function of time of flight in Bianchi I metrics, we present an independent derivation confirming the result by Saunders (1969). The present fit differs from earlier ones by Koivisto & Mota and by Campanelli et al. in that we use Saunders' formula, a larger sample of supernovae, Union 2 and JLA, and we use the general Bianchi I metric with three distinct eigenvalues.