Optimizing blind reconstruction of CMB B-modes for future experiments

TL;DR

This paper introduces two advanced NILC-based methods, MC-NILC and ocMILC, to improve the removal of Galactic foreground contamination in CMB B-mode polarization measurements for future experiments.

Contribution

The paper presents novel extensions of NILC, namely MC-NILC and ocMILC, designed to enhance foreground subtraction in CMB B-mode reconstruction.

Findings

MC-NILC and ocMILC reduce residual Galactic contamination.

New methods outperform previous NILC in complex foreground scenarios.

Improved foreground removal leads to more accurate CMB B-mode detection.

Abstract

The detection of primordial polarization $B$ modes of the Cosmic Microwave Background (CMB) requires exquisite control of Galactic foreground contamination. The Needlet Internal Linear Combination (NILC) method has proven effective in reconstructing CMB $B$ modes without suffering from mis-modeling errors of Galactic emission. However, with the most complex foreground models, residual Galactic contamination from NILC is proved to bias, especially at large angular scales, the recovered CMB $B$ modes from simulated data of future CMB experiments. We therefore present two new extensions of NILC, Multi-Clustering NILC (MC-NILC) and optimized constrained Moment ILC (ocMILC), which allow to enhance foreground subtraction in the reconstructed CMB signal.