Phase statistics of the WMAP 7 year data

TL;DR

This study analyzed WMAP 7-year data's phase statistics to detect non-Gaussianities, finding initial signals likely due to foreground contamination and constraining primordial non-Gaussianity parameters.

Contribution

It provides a comprehensive phase-based analysis of WMAP data, confirming non-Gaussianity is due to foregrounds and offering an updated constraint on f_NL.

Findings

Detected non-Gaussian signals at high-l in a_lm coefficients.

Foreground contamination explains observed non-Gaussianity.

Estimated f_NL=40 +/- 200, consistent with previous results.

Abstract

We performed a comprehensive statistical analysis using complex phases of the a_lm coefficients computed from the most recent data of the Wilkinson Microwave Anisotropy Probe (WMAP). Our aim was to confirm or constrain the presence of non-Gaussianities in the data. We found phase correlations - that suggest non-Gaussianity - at high-l in a_lm coefficients by applying various statistical tests. Most of all, we detected a non-Gaussian signal reaching a significance of 4.7 sigma using random walk statistics and simulations. However, our conclusion is that the non-Gaussian behavior is due to contamination from galactic foregrounds that show up in small scales only. When masked out the contaminated regions, we found no significant non-Gaussianity. Furthermore, we constrained the f_NL parameter using CMB simulations that mimic primordial non-Gaussianity. Our estimate is f_NL=40 +/- 200, in agreement with previous measurements and inflationary expectations.