The properties, origin and evolution of stellar clusters in galaxy simulations and observations

Published onlineWe investigate the properties and evolution of star particles in two simulations of isolated spiral galaxies, and two galaxies from cosmological simulations. Unlike previous numerical work, where typically each star particle represents one ‘cluster’, for the isolated galaxies we are able to model features we term ‘clusters’ with groups of particles. We compute the spatial distribution of stars with different ages, and cluster mass distributions, comparing our findings with observations including the recent LEGUS survey. We find that spiral structure tends to be present in older (100s Myr) stars and clusters in the simulations compared to the observations. This likely reflects differences in the numbers of stars or clusters, the strength of spiral arms, and whether the clusters are allowed to evolve. Where we model clusters with multiple particles, we are able to study their evolution. The evolution of simulated clusters tends to follow that of their natal gas clouds. Massive, dense, long-lived clouds host massive clusters, whilst short-lived clouds host smaller clusters which readily disperse. Most clusters appear to disperse fairly quickly, in basic agreement with observational findings. We note that embedded clusters may be less inclined to disperse in simulations in a galactic environment with continuous accretion of gas on to the clouds than isolated clouds and correspondingly, massive young clusters which are no longer associated with gas tend not to occur in the simulations. Caveats of our models include that the cluster densities are lower than realistic clusters, and the simplistic implementation of stellar feedback.We thank the referee for a useful report. The calculations for this paper were performed primarily on the DiRAC machine ‘Complexity’, as well as the supercomputer at Exeter, which is jointly funded by STFC, the Large Facilities Capital Fund of BIS, and the University of Exeter. We would like to thank Michele Fumagalli for work putting together the LEGUS cluster catalogues. CLD and CGF acknowledge funding from the European Research Council for the FP7 ERC starting grant project LOCALSTAR. CGF thanks Ben Thompson for performing data reduction. DG kindly acknowledges financial support by the German Research Foundation (DFG) through grant GO 1659/3-2. Figures in this paper were produced using splash (Price 2007)