X-ray morphological estimators for galaxy clusters

TL;DR

This paper evaluates various X-ray morphological parameters to classify galaxy clusters as regular or disturbed, introducing a new combined parameter that enhances classification accuracy.

Contribution

It reviews and tests multiple morphological estimators on simulated data, identifying the most effective indicators and proposing a new combined parameter for better classification.

Findings

Third-order power ratio, asymmetry, and concentration are top classifiers.

Combining indicators yields near-perfect purity in classification.

The new parameter M outperforms individual estimators.

Abstract

The classification of galaxy clusters according to their X-ray appearance is a powerful tool to discriminate between regular clusters (associated to relaxed objects) and disturbed ones (linked to dynamically active systems). The compilation of the two subsamples is a necessary step both for cosmological studies - oriented towards spherical and virialized systems- and for astrophysical investigations - focused on phenomena typically present in highly disturbed clusters such as turbulence, particle re-acceleration, magneto-astrophysics . In this paper, we review several morphological parameters: asymmetry and fluctuation of the X-ray surface brightness, hardness ratios, X-ray surface-brightness concentration, centroid shift, and third-order power ratio. We test them against 60 Chandra-like images obtained from hydrodynamical simulations through the X-ray Map Simulator and visually classified as regular and disturbed. The best performances are registered when the parameters are computed using the largest possible region (either within R500 or 1000 kpc). The best indicators are the third-order power ratio, the asymmetry parameter, and the X-ray-surface-brightness concentration. All their combinations offer an efficient way to distinguish between the two morphological classes achieving values of purity extremely close to 1. A new parameter, M, is defined. It combines the strengths of the aforementioned indicators and, therefore, resulted to be the most effective parameter analyzed.