Primordial trispectrum from kSZ tomography

TL;DR

This paper proposes a method using kSZ tomography to measure primordial non-Gaussianity, specifically the trispectrum amplitude, with forecasts showing significant sensitivity improvements over existing galaxy survey methods.

Contribution

It introduces a novel approach combining CMB and galaxy data to probe the primordial trispectrum, achieving sample variance cancellation and enhanced sensitivity.

Findings

Forecasted sensitivity of σ_{τ_NL} ≈ 1.5 for CMB-S4 and VRO

Achieves 10-fold improvement in trispectrum measurement over galaxy surveys alone

Demonstrates the importance of high galaxy density and large survey volume

Abstract

The kinetic Sunyaev Zel'dovich effect is a secondary CMB temperature anisotropy that provides a powerful probe of the radial-velocity field of matter distributed across the Universe. This velocity field is reconstructed by combining high-resolution CMB measurements with galaxy survey data, and it provides an unbiased tracer of matter perturbations in the linear regime. In this paper, we show how this measurement can be used to probe primordial non-Gaussianity of the local type, particularly focusing on the trispectrum amplitude $\tau_{\rm NL}$, as may arise in a simple two-field inflation model that we provide by way of illustration. Cross-correlating the velocity-field-derived matter distribution with the biased large-scale galaxy density field allows one to measure the scale-dependent bias factor with sample variance cancellation. We forecast that a configuration corresponding to CMB-S4 and VRO results in a sensitivity of $\sigma_{f_{\rm NL}} \approx 0.59$ and $\sigma_{\tau_{\rm NL}} \approx 1.5$. These forecasts predict improvement factors of 10 and 195 for $\sigma_{f_{\rm NL}}$ and $\sigma_{\tau_{\rm NL}}$, respectively, over the sensitivity using VRO data alone, without internal sample variance cancellation. Similarly, for a configuration corresponding to DESI and SO, we forecast a sensitivity of $\sigma_{f_{\rm NL}} \approx 3.1$ and $\sigma_{\tau_{\rm NL}} \approx 69$, with improvement factors of 2 and 5, respectively, over the use of the DESI data-set in isolation. We find that a high galaxy number density and large survey volume considerably improve our ability to probe the amplitude of the primordial trispectrum for the multi-field model considered.