Delensing Galaxy Surveys

TL;DR

This paper investigates the impact of gravitational lensing displacements on galaxy images, demonstrating that delensing is feasible with large surveys and testing methods using shear and foreground galaxy distributions.

Contribution

It introduces methods to reconstruct original galaxy positions affected by lensing displacements using shear and foreground galaxy data.

Findings

Delensing is feasible with surveys covering hundreds of square degrees.

Displacement effects are redshift-dependent and can be corrected.

Foreground deflections impact galaxy-galaxy lensing mainly for LSST and Euclid.

Abstract

Weak gravitational lensing can cause displacements, magnification, rotation and shearing of the images of distant galaxies. Most studies focus on the shear and magnification effects since they are more easily observed. In this paper we focus on the effect of lensing displacements on wide field images. Galaxies at redshifts 0.5--1 are typically displaced by 1 arcminute, and the displacements are coherent over degree-size patches. However the displacement effect is redshift-dependent, so there is a visible relative shift between galaxies at different redshifts, even if they are close on the sky. We show that the reconstruction of the original galaxy position is now feasible with lensing surveys that cover many hundreds of square degrees. We test with simulations two approaches to "delensing": one uses shear measurements and the other uses the foreground galaxy distribution as a proxy for the mass. We also estimate the effect of foreground deflections on galaxy-galaxy lensing measurements and find it is relevant only for LSST and Euclid-era surveys.