Contamination of early-type galaxy alignments to galaxy lensing-CMB lensing cross-correlation

TL;DR

This paper reevaluates the contamination of galaxy intrinsic alignments on galaxy lensing-CMB lensing cross-correlation, finding a moderate impact from red galaxies and significant uncertainty at higher redshifts, which affects cosmological measurements.

Contribution

It provides updated estimates of intrinsic alignment contamination using recent observational constraints, highlighting the uncertainties at higher redshifts and the potential for substantial contamination.

Findings

Red galaxy contamination is approximately 10% with 3% uncertainty.

Blue galaxies are consistent with being unaligned but could add up to 9.5% contamination.

At high redshift (z>1.2), contamination could reach up to 60% under extreme assumptions.

Abstract

Galaxy shapes are subject to distortions due to the tidal field of the Universe. The cross-correlation of galaxy lensing with the lensing of the Cosmic Microwave Background (CMB) cannot easily be separated from the cross-correlation of galaxy intrinsic shapes with CMB lensing. Previous work suggested that the intrinsic alignment contamination can be $15\%$ of this cross-spectrum for the CFHT Stripe 82 (CS82) and Atacama Cosmology Telescope surveys. Here we re-examine these estimates using up-to-date observational constraints of intrinsic alignments at a redshift more similar to that of CS82 galaxies. We find a $\approx$ $10\%$ contamination of the cross-spectrum from red galaxies, with $\approx$ $3\%$ uncertainty due to uncertainties in the redshift distribution of source galaxies and the modelling of the spectral energy distribution. Blue galaxies are consistent with being unaligned, but could contaminate the cross-spectrum by an additional $9.5\%$ within current $95\%$ confidence levels. While our fiducial estimate of alignment contamination is similar to previous work, our work suggests that the relevance of alignments for CMB lensing-galaxy lensing cross-correlation remains largely unconstrained. Little information is currently available about alignments at $z>1.2$. We consider the upper limiting case where all $z>1.2$ galaxies are aligned with the same strength as low redshift luminous red galaxies, finding as much as $\approx$ $60\%$ contamination.