Primordial Non-Gaussianity and the NRAO VLA Sky Survey

TL;DR

This study investigates primordial non-Gaussianity of local type by constraining $f_{ m NL}$ through the NVSS auto-correlation function $w(\theta)$. The authors model how local non-Gaussianity induces a $1/k^2$ scale-dependent halo bias and incorporate it into the halo mass function, using CosmoMC to fit $f_{ m NL}$ and the minimum halo mass $M_{ m min}$ against external cosmological data. They find $f_{ m NL}=62\pm27$ ($1\sigma$) and $M_{ m min}=10^{12.47\pm0.26}\,h^{-1}M_\odot$, with $25<f_{ m NL}<117$ at 95% CL, indicating a non-Gaussian signal at about $3\sigma$; results are robust to bias-model variations and cross-checks, though subject to jackknife covariance limitations. The work demonstrates that the NVSS ACF can serve as a meaningful probe of primordial non-Gaussianity and motivates future wide-area radio surveys to tighten these constraints on scales of $\sim$100 Mpc.

Abstract

The NRAO VLA Sky Survey (NVSS) is the only dataset that allows an accurate determination of the auto-correlation function (ACF) on angular scales of several degrees for Active Galactic Nuclei (AGNs) at typical redshifts $z \simeq 1$. Surprisingly, the ACF is found to be positive on such large scales while, in the framework of the standard hierarchical clustering scenario with Gaussian primordial perturbations it should be negative for a redshift-independent effective halo mass of order of that found for optically-selected quasars. We show that a small primordial non-Gaussianity can add sufficient power on very large scales to account for the observed NVSS ACF. The best-fit value of the parameter $f_{\rm NL}$, quantifying the amplitude of primordial non-Gaussianity of local type is $f_{\rm NL}=62 \pm 27$ ($1\,σ$ error bar) and $25<f_{\rm NL}<117$ ($2\,σ$ confidence level), corresponding to a detection of non-Gaussianity significant at the $\sim 3\,σ$ confidence level. The minimal halo mass of NVSS sources is found to be $M_{\rm min}=10^{12.47\pm0.26}h^{-1}M_{\odot}$ ($1\,σ$) strikingly close to that found for optically selected quasars. We discuss caveats and possible physical and systematic effects that can impact on the results.

Figures (5)

• Figure 1: Observed ACF of NVSS catalog. Values are jackknife estimated. The black solid line is the best fit model of our non-Gaussian calculations, while the red dashed line refers to the Gaussian case. The vertical arrow marks the angular scale above which the theoretical Gaussian ACF becomes negative. (Here, negative values are not visible, due to their very small amplitudes. See Figure \ref{['fig:bias_one']} for details.)
• Figure 2: Effect of different M$_{\rm min}$ masses on the ACF: the zero-crossing angular scale of the ACF decreases with increasing $M_{\rm min}$ for $f_{\rm NL}=0$.
• Figure 3: Effects of non-Gaussianity on the auto-correlation power spectra (left panel) and on the ACF (right panel) for three different Gaussian bias models.
• Figure 4: Marginalized one-dimensional and two-dimensional distributions (1, 2$\,\sigma$ contours) of the minimal halo mass $M_{\rm min}$ and of the non-Gaussian parameter $f_{\rm NL}$.
• Figure 5: Dependence on the maximum separation $\theta$ of the error bars on the minimal halo mass $M_{\rm min}$ and on $f_{\rm NL}$, calculated with the Fisher matrix method (arbitrary normalization).