A Bayesian ILC method for CMB B-mode posterior estimation and reconstruction of primordial gravity wave signal

TL;DR

This paper introduces a Bayesian ILC method for reconstructing CMB B-mode signals and estimating primordial gravitational waves, effectively handling foreground contamination and noise bias, demonstrated through extensive simulations.

Contribution

It extends the Bayesian ILC approach with noise bias corrections, enabling accurate weak CMB B-mode reconstruction and joint posterior density estimation from simulated future CMB data.

Findings

Accurately reconstructs CMB B-mode signals over large sky areas.

Achieves unbiased detection of primordial gravitational wave signal with high significance.

Demonstrates effectiveness through 200 Monte Carlo simulations.

Abstract

The Cosmic Microwave Background (CMB) radiation B mode polarization signal contains the unique signature of primordial metric perturbations produced during the inflation. The separation of the weak CMB B-mode signal from strong foreground contamination in observed maps is a complex task, and proposed new generation low noise satellite missions compete with the weak signal level of this gravitational background. In this article, for the first time, we employ a foreground model-independent internal linear combination (ILC) method to reconstruct the CMB B mode signal using simulated observations over large angular scales of the sky of 6 frequency bands of future generation CMB mission Probe of Inflation and Cosmic Origins (PICO). We estimate the joint CMB B mode posterior density following the interleaving Gibbs steps of B mode angular power spectrum and cleaned map samples using the ILC method. We extend and improve the earlier reported Bayesian ILC method to analyze weak CMB B mode reconstruction by introducing noise bias corrections at two stages during the ILC weight estimation. By performing $200$ Monte Carlo simulations of the Bayesian ILC method, we find that our method can reconstruct the CMB signals and the joint posterior density accurately over large angular scales of the sky. We estimate Blackwell-Rao statistics of the marginal density of CMB B mode angular power spectrum and use them to estimate the joint density of scalar to tensor ratio $r$ and a lensing power spectrum amplitude $A^{\textrm{lens}}$. Using $200$ Monte Carlo simulations of the delensing approach, we find that our method can achieve an unbiased detection of the primordial gravitational wave signal $r$ with more than 8$\sigma$ significance for levels of $r \geqslant 0.01$.