Abstract We identify compact groups (CGs) of galaxies in an IllustrisTNG300 simulation using a friends-of-friends algorithm. Our approach is designed to be comparable to systematic CG searches based on spectroscopic surveys, while avoiding the conventional Hickson selection criteria, which can bias samples toward relatively low-density environments. We construct two CG catalogs: one based on a three-dimensional distance linking length of 73 kpc (i.e., 50 h −1 kpc), and another based on projected and radial linking lengths of 73 kpc and 1000 km s −1 . We refer to these as the position–position–position (PPP) and position–position–velocity (PPV) CG catalogs, respectively. The PPV catalog provides a direct analog to observed CG samples. At z = 0 in TNG300, we identify 383 PPP CGs and 1666 PPV CGs. A large fraction (∼80%) of PPV CGs are not physically compact systems but are contaminated by line-of-sight interlopers. We demonstrate that the scaling relation between total group stellar mass and velocity dispersion is an effective diagnostic for identifying false positives with line-of-sight interlopers. We further examine the large-scale environments of CGs and show that they reside in a wide range of densities, including the central regions of galaxy clusters. These CG catalogs provide a robust foundation for studying the formation and evolution of CGs in cosmological simulations.
Yoo et al. (Fri,) studied this question.