A numerical study of the effects of primordial non-Gaussianities on weak lensing statistics

TL;DR

This study uses N-body simulations to assess how primordial non-Gaussianities influence weak lensing statistics, finding effects are small but potentially detectable with large surveys and careful parameter analysis.

Contribution

It provides a detailed numerical analysis of primordial non-Gaussianity effects on weak lensing observables using realistic simulations, highlighting degeneracies with other cosmological parameters.

Findings

Primordial non-Gaussianity effects are a few percent for current constraints.

Maximum effect reaches 10-15% for extreme non-Gaussianity levels.

Degeneracy with $\sigma_8$ and $\Omega_m$ affects detection prospects.

Abstract

While usually cosmological initial conditions are assumed to be Gaussian, inflationary theories can predict a certain amount of primordial non-Gaussianity which can have an impact on the statistical properties of the lensing observables. In order to evaluate this effect, we build a large set of realistic maps of different lensing quantities starting from light-cones extracted from large dark-matter only N-body simulations with initial conditions corresponding to different levels of primordial local non-Gaussianity strength $f_{\rm NL}$. Considering various statistical quantities (PDF, power spectrum, shear in aperture, skewness and bispectrum) we find that the effect produced by the presence of primordial non-Gaussianity is relatively small, being of the order of few per cent for values of $|f_{\rm NL}|$ compatible with the present CMB constraints and reaching at most 10-15 per cent for the most extreme cases with $|f_{\rm NL}|=1000$. We also discuss the degeneracy of this effect with the uncertainties due to the power spectrum normalization $\sigma_8$ and matter density parameter $\Omega_{\rm m}$, finding that an error in the determination of $\sigma_8$ ($\Omega_{\rm m}$) of about 3 (10) per cent gives differences comparable with non-Gaussian models having $f_{\rm NL}=\pm 1000$. These results suggest that the possible presence of an amount of primordial non-Gaussianity corresponding to $|f_{\rm NL}|=100$ is not hampering a robust determination of the main cosmological parameters in present and future weak lensing surveys, while a positive detection of deviations from the Gaussian hypothesis is possible only breaking the degeneracy with other cosmological parameters and using data from deep surveys covering a large fraction of the sky.