Probing CMB Secondary Anisotropies through Minkowski Functionals

TL;DR

This paper introduces a novel method using Minkowski Functionals and skew-spectra to analyze secondary non-Gaussian anisotropies in the CMB, enhancing the separation of different effects like ISW, tSZ, and lensing.

Contribution

It develops estimators for skew-spectra in the presence of observational masks and provides covariance expressions, improving the analysis of non-Gaussian signatures in CMB maps.

Findings

High signal-to-noise ratios for Planck and EPIC experiments.

Skew-spectra are correlated, especially at higher multipoles.

Lensing-induced skew-spectra have lower S/N compared to tSZ maps.

Abstract

Secondary contributions to the anisotropy of the Cosmic Microwave Background (CMB), such as the integrated Sachs-Wolfe (ISW) effect, the thermal Sunyaev-Zel'dovich effect (tSZ), and the effect of gravitational lensing, have distinctive non-Gaussian signatures, and full descriptions therefore require information beyond that contained in their power spectra. In this paper we use the recently introduced skew-spectra associated with the Minkowski Functionals (MF) to probe the topology of CMB maps to probe the secondary non-Gaussianity as a function of beam-smoothing in order to separate various contributions. We devise estimators for these spectra in the presence of a realistic observational masks and present expressions for their covariance as a function of instrumental noise. Specific results are derived for the mixed ISW-lensing and tSZ-lensing bispectra as well as contamination due to point sources for noise levels that correspond to the Planck (143 GHz channel) and EPIC (150 GHz channel) experiments. The cumulative signal to noise ration $S/N$ for one-point generalized skewness-parameters can reach an order of ${\cal O}(10)$ for Planck and two orders of magnitude higher for EPIC, i.e. ${\cal O}(10^3)$. We also find that these three spectra skew-spectra are correlated, having correlation coefficients $r \sim 0.5-1.0$; higher $l$ modes are more strongly correlated. Though the values of $S/N$ increase with decreasing noise, the triplets of skew-spectra that determine the MFs bcome more correlated; the $S/N$ ratios of lensing-induced skew-spectra are smaller compared to that of a frequency-cleaned tSZ map.