Galaxy Modeling with Compound Elliptical Shapelets

TL;DR

This paper introduces a new compound shapelet basis and convolution relation to improve galaxy modeling, especially for highly elliptical or high Sersic index galaxies, reducing underfitting bias in weak-lensing analyses.

Contribution

It proposes a novel multi-scale shapelet basis and convolution relation that better represent complex galaxy shapes, addressing limitations of standard shapelet methods.

Findings

Demonstrates improved galaxy modeling with the new basis

Reduces underfitting bias in weak-lensing measurements

Provides proof-of-concept with nearby galaxy sample

Abstract

Gauss-Hermite and Gauss-Laguerre ("shapelet") decompositions of images have become important tools in galaxy modeling, particularly for the purpose of extracting ellipticity and morphological information from astronomical data. However, the standard shapelet basis functions cannot compactly represent galaxies with high ellipticity or large Sersic index, and the resulting underfitting bias has been shown to present a serious challenge for weak-lensing methods based on shapelets. We present here a new convolution relation and a compound "multi-scale" shapelet basis to address these problems, and provide a proof-of-concept demonstration using a small sample of nearby galaxies.