http://arxiv.org/abs/1610.03351
Next-generation galaxy surveys will increasingly rely on the galaxy bispectrum to improve cosmological constraints, especially on primordial non-Gaussianity. A key theoretical requirement that remains to be developed is the analysis of general relativistic effects on the bispectrum, which arise from observing galaxies on the past lightcone. Here we compute for the first time all the local relativistic corrections to the bispectrum, from Doppler, gravitational potential and higher-order effects. For the galaxy bispectrum, the problem is much more complex than for the power spectrum, since we need the lightcone corrections at second order. Mode-coupling contributions at second order mean that relativistic corrections can be non-negligible at smaller scales than in the case of the power spectrum. In a primordial Gaussian universe, we show that the relativistic bispectrum for a moderately squeezed shape can differ from the Newtonian prediction by $\sim 30\%$ when the short modes are at the equality scale. For the equilateral shape, the difference is $\sim 20\%$ at gigaparsec scales. The relativistic corrections, if ignored in the analysis of observations, could therefore easily be mistaken for primordial non-Gaussianity. We conclude that for upcoming surveys which probe equality scales and beyond, these new relativistic signatures must be included for an accurate measurement of primordial non-Gaussianity.
O. Umeh, S. Jolicoeur, R. Maartens, et. al.
Wed, 12 Oct 16
59/64
Comments: 7 pages, 5 figures
You must be logged in to post a comment.