Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
What Globular Clusters Teach Us About The Assembly Of Their Host Galaxies Oleg Gnedin (University of Michigan) modeling formation The Galaxy – Halo Connection • In the first two days of this meeting, we talked about matching galaxies to halos, in the full range from faint galaxies to massive galaxy clusters • While the distribution of dark matter is approximately self-similar on different mass scales, the baryon physics is not: Inside galaxies, the gas has higher density and shorter cooling time; significant fraction of baryons is in stars • We can trace the assembly of galaxies by following their stellar components. These are important observational constraints because: Modeling star formation, and feedback of young stars and black holes on the interstellar medium, is complex (many processes and huge dynamic range, from > 10 Mpc to < 1 pc) Globular clusters are unique tracers of early star formation Spherical clusters of stars, from ∼104 M to 107 M Very high stellar density, up to 105 M pc-3 , half-mass radius 1-5 pc Remain self-gravitating for a Hubble time and are easy to pick up in photometric surveys Single-age stellar populations (can measure stellar IMF) Ages 8 – 13 Gyr, formation redshift > 1 GC metallicity reflects the conditions of galactic ISM at the time of formation: window into early star formation Li & OG 2014 Amazingly simple connection to halos fraction of galaxy mass in the globular cluster system is similar in halos of all type and environment over 5 orders of magnitude in mass group environment Spitler & Forbes (2009) Georgiev at al. (2010) Harris et al. (2013) Hudson et al. (2014) single galaxy environment dwarfs are problem MGC ≈ (1-5)×10-3 M* ≈ 10-4 Mhost In contrast, fairly complex metallicity distributions blue red [Fe/H]≈−0.5 [Fe/H]≈−1.5 Peng et al. 2006 – HST Virgo Cluster Survey Li & OG 2014 Color for old stellar population is a good proxy for metallicity Metallicity distribution of GCs in most galaxies is multimodal Each metallicity group tells us about different episodes of cluster formation Build-up of the globular cluster system from gas-rich mergers and accretion of galaxies: begin with cosmological simulations of halo formation supplement halos with cold gas mass based on observations use MGC - Mgas relation from hydro simulations metallicity from observed M* - Z relation for host galaxies, including evolution with time Muratov & OG 2010 Li & OG 2014 Hui Li Formation of massive star clusters may be triggered/enhanced by major mergers of gas-rich galaxies • Star clusters form in giant molecular clouds (GMC) • Higher density of globular clusters requires higher external pressure on GMC • Interstellar medium is over-pressured during galaxy mergers • Extreme masses of proto-globular GMC require very gas-rich galaxies In addition, accreted satellites bring their populations of GCs outer GCs in M31 are associated with tidal streams (D. Mackey) Basic idea: • Globular clusters form in galaxies throughout cosmic time • Galaxies merge into bigger galaxies • Their globular cluster systems combine • Can we reconstruct the assembly history of the host galaxy from observations of its globular cluster system? Need better understanding of the conditions leading to the formation of massive star clusters Important for modeling galaxy formation in general: young star clusters are dominant components of active star formation Adamo et al. 2015 Fraction of all young stars contained in massive star clusters increases with the intensity of star formation, up to 50-60% Initial Mass Function of Young Clusters is almost invariant z≈5 Cosmological simulation of a Milky Way sized-galaxy (Li et al. 2016): • MF is a power law as observed for young star clusters • Unless star formation is too disruptive z ≈ 3.3 Massive clusters form within galaxy disks, but the disks are perturbed by frequent mergers and accretion of satellites Simulations with the ART code. Novel treatment of subgrid-scale turbulence (see poster by Kravtsov) z≈4 Feedback of own stars terminates star formation within a few Myr and sets the final cluster mass ρ > 103 cm-3 εff = 10-100% also: ρ > 10 cm-3 εff from SGS turbulence model (Semenov et al. 2016) • spatial resolution 5 pc at z=5 • typical timestep 103 yr – resolved cluster growth • clusters grow from gas mass inside fixed-sized sphere of 10 pc (typical GMC core/clump) In the simulations, cluster mass function remains a stable power-law Different epochs within the same central galaxy Different galaxies at the same epoch (z ≈ 3.3) Cluster formation efficiency fraction of total mass of young stars contained in star clusters more massive than 104 M observations --- central galaxy --- first satellite --- other satellites Maximum cluster mass scales with the star formation rate Do mergers trigger cluster formation? Maybe (Mi-Mi-1)/Mi M2/M1 no merger major merger Cluster MF is truncated between mergers Summary • Globular clusters could be used to trace major episodes of mass assembly of their host galaxies • Red clusters are due to intermediate-z gas-rich mergers, blue clusters are due to early mergers and later massive mergers • At the time of formation, GC mass function is a power law, both in observations and galaxy formation simulations • Mergers of gas-rich galaxies may trigger/enhance formation rate of massive star clusters (M > 105 M)