CMB B-mode non-Gaussianity: optimal bispectrum estimator and Fisher forecasts

TL;DR

This paper develops an optimal bispectrum estimator for B-mode non-Gaussianity in the CMB, enabling tests of inflation models through Fisher forecasts for upcoming experiments.

Contribution

It introduces a generalized KSW bispectrum estimator for B-mode non-Gaussianity and provides Fisher forecasts for future CMB experiments.

Findings

Estimator retains numerical efficiency for high-resolution data

Forecasts show sensitivity to experimental parameters

Contaminant effects are discussed

Abstract

Upcoming cosmic microwave background (CMB) data can be used to explore harmonic 3-point functions that involve the B-mode component of the CMB polarization signal. We focus on bispectra describing the non-Gaussian correlation of the B-mode field and the CMB temperature anisotropies (T) and/or E-mode polarization, i.e. <TTB>, <EEB>, and <TEB>. Such bispectra probe violations of the tensor consistency relation: the model-independent behavior of cosmological correlation functions that involve a large-wavelength tensor mode (gravitational wave). An observed violation of the tensor consistency relation would exclude a large number of inflation models. We describe a generalization of the Komatsu-Spergel-Wandelt (KSW) bispectrum estimator that allows statistical inference on this type of primordial non-Gaussianity with data of the CMB temperature and polarization anisotropies. The generalized estimator shares its statistical properties with the existing KSW estimator and retains the favorable numerical scaling with angular resolution. In this paper we derive the estimator and present a set of Fisher forecasts. We show how the forecasts scale with various experimental parameters such as lower and upper angular band-limit, relevant for e.g. the upcoming ground-based Simons Observatory experiment and proposed LiteBIRD satellite experiment. We comment on possible contaminants due to secondary cosmological and astrophysical sources.