paper

Polarization as an indicator of intrinsic alignment in radio weak lensing

arXiv:1005.1926 · doi:10.1111/j.1365-2966.2010.17583.x

Abstract

We propose a new technique for weak gravitational lensing in the radio band making use of polarization information. Since the orientation of a galaxy's polarized emission is both unaffected by lensing and is related to the galaxy's intrinsic orientation, it effectively provides information on the unlensed galaxy position angle. We derive a new weak lensing estimator which exploits this effect and makes full use of both the observed galaxy shapes and the estimates of the intrinsic position angles as provided by polarization. Our method has the potential to both reduce the effects of shot noise, and to reduce to negligible levels, in a model-independent way, all effects of intrinsic galaxy alignments. We test our technique on simulated weak lensing skies, including an intrinsic alignment contaminant consistent with recent observations, in three overlapping redshift bins. Adopting a standard weak lensing analysis and ignoring intrinsic alignments results in biases of 5-10% in the recovered power spectra and cosmological parameters. Applying our new estimator to one tenth the number of galaxies used for the standard case, we recover both power spectra and the input cosmology with similar precision as compared to the standard case and with negligible residual bias, even in the presence of a substantial (astrophysical) scatter in the relationship between the observed orientation of the polarized emission and the intrinsic orientation. Assuming a reasonable polarization fraction for star-forming galaxies, and no cosmological conspiracy in the relationship between polarization direction and intrinsic morphology, our estimator should prove a valuable tool for weak lensing analyses of forthcoming radio surveys, in particular, deep wide field surveys with e-MERLIN, MeerKAT and ASKAP and ultimately, definitive radio lensing surveys with the SKA.

18 pages, 10 figures, submitted to MNRAS

References in corpus (10)

Cited by in corpus (30)