Non-Gaussianity of secondary anisotropies from ACTPol and Planck

TL;DR

This paper analyzes the non-Gaussian features of secondary anisotropies in the cosmic microwave background using ACTPol and Planck data, detecting signals from radio galaxies, tSZ effect, and dusty galaxies, and introduces a new method for covariance calculation.

Contribution

It presents a novel approach to quantify non-Gaussian contributions to bispectrum covariances and demonstrates the potential of bispectrum measurements for cosmological parameter constraints.

Findings

Strong detection of radio galaxy non-Gaussianity (>5σ)

Hints of non-Gaussianity from tSZ and DSFGs

Demonstration of bispectrum utility for constraining cosmology

Abstract

Most secondary sources of cosmic microwave background anisotropy (radio sources, dusty galaxies -DSFGs-, thermal Sunyaev Zel'dovich -tSZ- distortions from hot gas, and gravitational lensing) are highly non-Gaussian. Statistics beyond the power spectrum are therefore potentially important sources of information about the physics of these processes. We combine data from the Atacama Cosmology Telescope and with data from the Planck satellite (only using Planck data in the overlapping region) to constrain the amplitudes of a set of theoretical bispectrum templates from the tSZ effect, DSFGs, gravitational lensing, and radio galaxies. We make a strong detection of radio galaxies (>5$\sigma$) and have hints of non-Gaussianity arising from the tSZ effect, DSFGs, from cross-correlations between the tSZ effect and DSFGs and from cross-correlations among the tSZ effect, DSFGs and radio galaxies. These results suggest that the same halos host radio sources, DSFGs, and have tSZ signal. We present a new method to calculate the non-Gaussian contributions to the template covariances. Using this method we find significant non-Gaussian contributions to the variance and covariance of our templates, with templates involving the tSZ effect most effected. Strong degeneracies exist between the various sources at the current noise levels. In light of these degeneracies, combined with theoretical uncertainty in the templates, these results are a demonstration of this technique. With these caveats, we demonstrate the utility of future bispectrum measurements by using the tSZ bispectrum measurement to constrain a combination of the amplitude of matter fluctuations and the matter density to be $\sigma_8 \Omega_m^{0.17}=0.65^{+0.05}_{-0.06}$. Improvements in signal to noise from upcoming observations will enable the separation of bispectrum components and robust constraints on cosmological parameters.