Analysis of NILC performance on B-modes data of sub-orbital experiments

TL;DR

This paper evaluates the effectiveness of NILC component separation on sub-orbital CMB B-mode data, addressing challenges like E-B leakage and demonstrating its potential for future experiments to detect primordial gravitational waves.

Contribution

It introduces extensions to NILC processing, such as iterative B-decomposition and inpainting, to improve B-mode recovery from partial-sky sub-orbital data.

Findings

E-B leakage correction techniques are effective for these datasets.

NILC can recover B-mode signals with minimal bias after processing.

Foreground subtraction with NILC achieves sensitivities compatible with experiment goals.

Abstract

The observation of primordial B-modes in the Cosmic Microwave Background (CMB) represents the main scientific goal of most of the future CMB experiments. This signal is predicted to be much lower than polarised Galactic emission (foregrounds) in any region of the sky pointing to the need for effective components separation methods, such as the Needlet-ILC (NILC). In this work, we explore the possibility of employing NILC for B-mode maps reconstructed from partial-sky data of sub-orbital experiments, addressing the complications that such an application yields: E-B leakage, needlet filtering and beam convolution. We consider two complementary simulated datasets from future experiments: the balloon-borne SWIPE telescope of the Large Scale Polarization Explorer, which targets the observation of both reionisation and recombination peaks of the primordial B-mode angular power spectrum, and the ground-based Small Aperture Telescope of Simons Observatory, which is designed to observe only the recombination bump. We assess the performance of two alternative techniques to correct for the CMB E-B leakage: the recycling technique (Liu et al. 2019) and the ZB method (Zhao & Baskaran 2010). We find that they both reduce the E-B leakage residuals at a negligible level given the sensitivity of the considered experiments, except for the recycling method in the SWIPE patch at $\ell < 20$. Thus, we implement two extensions of the pipeline, the iterative B-decomposition and the diffusive inpainting, which enable us to recover the input CMB B-mode power for $\ell \geq 5$. We demonstrate that needlet filtering and beam convolution do not affect the B-mode reconstruction. Finally, with an appropriate masking strategy, we find that NILC foregrounds subtraction allows to achieve sensitivities for the tensor-to-scalar ratio compatible to the targets of the considered CMB experiments.