Speaker
Description
Current searches for non-Gaussianity are mostly limited to higher-order correction functions of scalar perturbations. In the meantime, primordial B-mode detection is one of the main goals of next-generation CMB experiments. Bispectra involving B-modes, i.e., $\langle TTB \rangle, \langle EEB \rangle$, and $\langle TEB \rangle$ are of great importance for two reasons:
First, tensor perturbations source B modes while scalars do not, searching for bispectra involving B modes could put stronger constraints on tensor non-Gaussianity, which is predicted in a number of compelling theoretical models, i.e., inflation models involving axion-gauge field interactions, primordial magnetic field as a mechanism to generate tensor non-Gaussianity.
Second, current constraints on primordial tensor perturbations from T and E modes are limited by cosmic variance and B-mode observations are not yet cosmic-variance limited. Therefore, including future B-mode data from upcoming CMB experiments will improve our constraints on bispectra involving primordial tensor perturbations.
In this talk we describe a generalization of the Komatsu-Spergel-Wandelt(KSW) bispectrum estimator which shares its statistical properties with the existing KSW Estimator and maintains the favorable numerical scaling with angular resolution. We apply this estimator to SO and advanced SO data and present a set of preliminary constraints on $f_{NL}$ for bispectra involving tensor perturbation, i.e., $\langle \zeta \zeta h\rangle$, $\langle \zeta h h\rangle$, $\langle h h h\rangle$.