Preview

PDF, English Download (8MB) | Terms of use

Abstract

Is gravitational flexion reliably measurable in realistic observational conditions? To address this question we undertook an exhaustive investigation into the problem of estimating flexion. By means of synthetic data, we tested the ability of different methods to recover the shape of faint background galaxies with the level of detail necessary to account for the spin-1 and spin-3 distortions caused by flexion. Three different methods for weak lensing measurements have been extended in order to measure the high-order moments of surface brightness necessary to quantify the flexion-induced deformations. We observed that, for the two methods which exploit a weight function to control the noise and afterwards correct for the impact of this procedure, a successful outcome depends upon numerous assumptions on the properties of the objects investigated, making these techniques less appealing for practical applications. The third method we tested employs a principal component analysis algorithm to de-noise the images. Even though, in the most strict observational conditions, we could not achieve a precise measurement of the high-order distortions, this method stood out as a promising technique for shape measurements in weak lensing applications. Our analysis shows that a deeper understanding of the impact of pixel noise on the flexion estimators is required before measurements of flexion in real data can be carried out and, finally, used to exploit the vast potential of gravitational flexion.