CMB lensing reconstruction with point source masks

TL;DR

This paper demonstrates that masking point sources in CMB maps does not significantly impair the reconstruction of weak lensing convergence using a real-space quadratic estimator, ensuring reliable lensing analysis despite incomplete sky coverage.

Contribution

It introduces a real-space quadratic estimator approach that effectively reconstructs CMB lensing signals even with point source masks, showing robustness against masking artifacts.

Findings

Lensing signal can be recovered without significant power loss.

Masking defected pixels does not introduce spurious correlations.

Real-space quadratic estimator is effective with masked CMB maps.

Abstract

An incomplete sky coverage poses difficulties in the extraction of the weak lensing information from the CMB. We test the reconstruction of the weak lensing convergence from CMB maps to which masks of point sources have been applied. We use the quadratic estimator with a kernel with finite support acting in real space for a Planck simulation. We recover the lensing signal without significant loss of power or addition of spurious correlations, thus showing that masking defected pixels does not affect the reconstruction of the weak lensing convergence in real space.