EDP Sciences
Free Access
Volume 474, Number 3, November II 2007
Page(s) 717 - 729
Section Cosmology (including clusters of galaxies)
DOI https://doi.org/10.1051/0004-6361:20077998

A&A 474, 717-729 (2007)
DOI: 10.1051/0004-6361:20077998

(An)isotropy of the Hubble diagram: comparing hemispheres

D. J. Schwarz and B. Weinhorst

Fakultät für Physik, Postfach 100131, Universität Bielefeld, 33501 Bielefeld, Germany
    e-mail: [dschwarz;bwein]@physik.uni-bielefeld.de

(Received 1 June 2007 / Accepted 28 August 2007)

Aims.We test the isotropy of the Hubble diagram. At small redshifts, this is possible without assumptions on the cosmic inventory and provides a fundamental test of the cosmological principle. At higher redshift we check for the self-consistency of the $\Lambda$CDM model.
Methods.At small redshifts, we use public supernovae (SNe) Ia data to determine the deceleration parameter q0 and the SN calibration on opposite hemispheres. For the complete data sets we fit $\Omega_{\rm M}$ and the SN calibration on opposite hemispheres.
Results.A statistically significant anisotropy of the Hubble diagram at redshifts z < 0.2 is discovered (>95%C.L.). While data from the North Galactic hemisphere favour the accelerated expansion of the Universe, data from the South Galactic hemisphere are not conclusive. The hemispheric asymmetry is maximal toward a direction close to the equatorial poles. The discrepancy between the equatorial North and South hemispheres shows up in the SN calibration. For the $\Lambda$CDM model fitted to all available SNe, we find the same asymmetry.
Conclusions.The alignment of discrepancies between hemispheric Hubble diagrams with the equatorial frame seems to point toward a systematic error in the SN search, observation, analysis or data reduction. We also find that our model independent test cannot exclude the case of the deceleration of the expansion at a statistically significant level.

Key words: cosmology: observations -- large-scale structure of Universe -- supernovae: general

© ESO 2007