DS+: a method for the identification of cluster substructures

TL;DR

DS+ is a new, reliable method for identifying and characterizing substructures in galaxy clusters using positions and velocities, tested on simulations and real data, improving understanding of cluster dynamics and dark matter.

Contribution

The paper introduces DS+, an improved method for detecting cluster substructures based on galaxy positions and velocities, extending the traditional Dressler & Shectman approach.

Findings

DS+ identifies ~80% of real group galaxies as substructure members.

At least 60% of galaxies in detected substructures are true group members.

Application to the Bullet cluster confirms known collision features.

Abstract

The study of cluster substructures is important for the determination of the cluster dynamical status, assembly history, and the evolution of cluster galaxies, and it allows to set of constraints on the nature of dark matter and cosmological parameters. We present and test DS+, a new method for the identification and characterization of group-sized substructures in clusters. Our new method is based on the projected positions and line-of-sight velocities of cluster galaxies, and it is an improvement and extension of the traditional method of Dressler & Shectman (1988). We test it on cluster-size cosmological halos extracted from the IllustrisTNG simulations, with virial masses $\rm{14 \lesssim \log (M_{200}/M_{\odot}) \lesssim 14.6}$, that contain on average $\sim 190$ galaxies. We also present an application of our method on a real data set, the Bullet cluster. DS+ is able to identify $\sim 80\%$ of real group galaxies as members of substructures, and at least 60\% of the galaxies assigned to substructures belong to real groups. The physical properties of the real groups are significantly correlated with those of the corresponding detected substructures, albeit with significant scatter, and overestimated on average. Application of the DS+ method to the Bullet cluster confirms the presence and main properties of the high-speed collision and identifies other substructures along the main cluster axis. DS+ proves to be a reliable method for the identification of substructures in clusters. The method is made freely available to the community as a Python code.