Weak lensing using only galaxy position angles

• Published in: arXiv.org 2014-09
• Main Authors: Whittaker, Lee, Brown, Michael L, Battye, Richard
• Format: Article
• Language: English
• Subjects: Algorithms; Computer simulation; Convergence; Correlation analysis; Ellipticity; Estimates; Galaxies; Galaxy distribution; Normal distribution; Position measurement; Shear; Simulation; Stars & galaxies; Statistical analysis
• ISSN: 2331-8422
• DOI: 10.48550/arxiv.1311.3926

We develop a method for performing a weak lensing analysis using only measurements of galaxy position angles. By analysing the statistical properties of the galaxy orientations given a known intrinsic ellipticity distribution, we show that it is possible to obtain estimates of the shear by minimizing a \(\chi^2\) statistic. The method is demonstrated using simulations where the components of the intrinsic ellipticity are taken to be Gaussian distributed. Uncertainties in the position angle measurements introduce a bias into the shear estimates which can be reduced to negligible levels by introducing a correction term into the formalism. We generalize our approach by developing an algorithm to obtain direct shear estimators given any azimuthally symmetric intrinsic ellipticity distribution. We introduce a method of measuring the position angles of the galaxies from noisy pixelized images, and propose a method to correct for biases which arise due to pixelization and correlations between measurement errors and galaxy ellipticities. We also develop a method to constrain the sample of galaxies used to obtain an estimate of the intrinsic ellipticity distribution such that fractional biases in the resulting shear estimates are below a given threshold value. We demonstrate the angle only method by applying it to simulations where the ellipticities are taken to follow a log-normal distribution. We compare the performance of the position angle only method with the standard method based on full ellipticity measurements by reconstructing lensing convergence maps from both numerical simulations and from the CFHTLenS data. We find that the difference between the convergence maps reconstructed using the two methods is consistent with noise.