Mitigating the effects of undersampling in weak lensing shear estimation with metacalibration

TL;DR

This paper examines how undersampling affects metacalibration-based shear measurements in weak lensing, identifies aliasing biases, and proposes a shape measurement method to meet space mission accuracy requirements.

Contribution

It demonstrates that using wider Gaussian weight functions in shape measurements reduces aliasing bias in undersampled images, making metacalibration viable for space missions.

Findings

Aliasing bias can reach 0.01 for Euclid and larger for Roman.

Wider Gaussian weights mitigate aliasing bias effectively.

Method remains robust across different PSF conditions.

Abstract

Metacalibration is a state-of-the-art technique for measuring weak gravitational lensing shear from well-sampled galaxy images. We investigate the accuracy of shear measured with metacalibration from fitting elliptical Gaussians to undersampled galaxy images. In this case, metacalibration introduces aliasing effects leading to an ensemble multiplicative shear bias about 0.01 for Euclid, and even larger for the Roman Space Telescope, well exceeding the missions' requirements. We find that this aliasing bias can be mitigated by computing shapes from weighted moments with wider Gaussians as weight functions, thereby trading bias for a slight increase in variance of the measurements. We show that this approach is robust to the point-spread function in consideration and meets the stringent requirements of Euclid for galaxies with moderate to high signal-to-noise ratios. We therefore advocate metacalibration as a viable shear measurement option for weak lensing from upcoming space missions.