The AIMSS Project II: Dynamical-to-Stellar Mass Ratios Across the Star Cluster - Galaxy Divide

TL;DR

This study investigates the high dynamical-to-stellar mass ratios in ultra compact dwarfs and compact ellipticals, proposing tidal stripping as a key process influencing their properties, supported by simulations and observational data.

Contribution

It introduces tidal stripping as a plausible explanation for elevated mass ratios in UCDs, supported by simulations matching observed trends across different stellar masses.

Findings

UCDs have higher mass ratios than globular clusters.

cEs at high masses have mass ratios near unity.

Tidal stripping simulations can reproduce observed mass ratio variations.

Abstract

The previously clear division between small galaxies and massive star clusters is now occupied by objects called ultra compact dwarfs (UCDs) and compact ellipticals (cEs). Here we combine a sample of UCDs and cEs with velocity dispersions from the AIMSS project with literature data to explore their dynamical-to-stellar mass ratios. We confirm that the mass ratios of many UCDs in the stellar mass range 10$^6$ -- 10$^9$ M$_{\odot}$ are systematically higher than those for globular clusters which have mass ratios near unity. However, at the very highest masses in our sample, i.e. 10$^9$ -- 10$^{10}$ M$_{\odot}$, we find that cE galaxies also have mass ratios of close to unity, indicating their central regions are mostly composed of stars. Suggested explanations for the elevated mass ratios of UCDs have included a variable IMF, a central black hole, and the presence of dark matter. Here we present another possible explanation, i.e. tidal stripping. Under various assumptions, we find that the apparent variation in the mass ratio with stellar mass and stellar density can be qualitatively reproduced by published tidal stripping simulations of a dwarf elliptical galaxy. In the early stages of the stripping process the galaxy is unlikely to be in virial equilibrium. At late stages, the final remnant resembles the properties of $\sim$10$^7$ M$_{\odot}$ UCDs. Finally, we discuss the need for more detailed realistic modelling of tidal stripping over a wider range of parameter space, and observations to further test the stripping hypothesis.