Joint Minkowski Functionals and Bispectrum Constraints on Non-Gaussianity in the CMB

TL;DR

This paper combines bispectrum and Minkowski functional analyses of CMB data to improve constraints on primordial non-Gaussianity, revealing a positive correlation and tighter bounds than individual methods.

Contribution

First to evaluate the covariance between bispectrum and Minkowski functionals for non-Gaussianity constraints, enabling their statistically combined analysis.

Findings

Estimators are positively correlated with r ~ 0.3.

Combined analysis yields fnl=37±28, reducing error by ~20%.

Method improves constraints on primordial non-Gaussianity.

Abstract

Two of the most commonly used tools to constrain the primordial non-Gaussianity are the bispectrum and the Minkowski functionals of CMB temperature anisotropies. These two measures of non-Gaussianity in principle provide distinct (though correlated) information, but in the past constraints from them have only been loosely compared, and not statistically combined. In this work we evaluate, for the first time, the covariance matrix between the local non-Gaussianity coefficient fnl estimated through the bispectrum and Minkowski functionals. We find that the estimators are positively correlated, with corerlation coefficient r ~ 0.3. Using the WMAP7 data to combine the two measures and accounting for the point-source systematics, we find the combined constraint fnl=37+/-28, which has a ~20% smaller error than either of the individual constraints.