The formation and evolution of globular cluster systems Joel Pfeffer, Nate Bastian (Liverpool, LJMU)
Introduction to stellar clusters Open clusters: few - 10 4 M few Myr - few Gyr solar metallicity disk of the Galaxy Pleiades Globular clusters (GCs): 10 4-10 6 M 10-13 Gyr low metallicity bulge/halo of the Galaxy M80
Globular clusters Populations around all normal (few 100) and massive galaxies (few 10,000 in brightest cluster galaxies; e.g. NGC 6166: 40,000 GCs, Harris et al. 2016) Some of the oldest luminous objects in the universe Was once thought GCs could only form in special conditions in the early Universe (e.g. Peebles & Dicke 1968; Fall & Rees 1985) Omega Centauri M80 47 Tucanae
Globular clusters as tools GCs have been used to: Determine structure of the Milky Way (Shapely 1918) Understand and calibrate stellar evolution and stellar populations (Eddington 1926) Constrain the formation and evolution of the Milky Way (Searle & Zinn 1978)... as well as other nearby galaxies (Brodie & Strader 2006, Kruijssen 2014) Place bounds on the age of the universe (Chaboyer et al. 1996)
A surprise with Hubble Globular clusters are still forming today: young massive clusters (YMCs) few 10 6 M 7 Myr Antennae colliding galaxies - HST similar density to GCs Whitmore et al. 1999
A surprise with Hubble Globular clusters are still forming today: young massive clusters (YMCs) NGC 34 10 7 M 100 Myr Schweizer & Seitzer 2007 Cabrera-Ziri et al. 2014
A surprise with Hubble Globular clusters are still forming today: young massive clusters (YMCs) NGC 7252 10 8 M 400 Myr 10 7 M 400 Myr Miller et al. 1997; Schwizer & Seitzer 1998; Maraston et al. 2004; Bastian et al. 2006 Sizes 10-20 pc
YMCs in our own Galaxy Westerlund 1 6 104 M 3.5 Myr (Clark & Negueruela 2004) 8 104 M few 100 Myr (Davies et al. 2011) GLIMPSE-C01
Fraction of U-band light from young clusters vs. Σ SFR Adamo et al. 2011 Specific luminosity: T L =100 L clusters /L galaxy Sample of quiescent spiral and star forming dwarf galaxies For increasing SFR density, the fraction of stars born in bound clusters is higher. Larsen 1999, 2000; Larsen & Richtler 1999, 2000
Fraction of star formation in clusters (Γ) vs. Σ SFR Γ = CFR/SFR Adamo et al. 2015 Young samples (<10 Myr) suffer from contamination Old samples suffer from disruption effects Between 1-50% of star formation happens in clusters Γ determined by gas pressure (Kruijssen 2012) (higher pressure higher star formation efficiencies more bound star formation) Goddard et al. 2010; Silva-Villa & Larsen 2011; Cook et al. 2012
Initial cluster mass function (ICMF) Power law index α = 2 follows from hierarchical growth (Elmegreen & Falgarone 1996; Fujii & Portegies Zwart 2015) Maximum mass scale increases with gas pressure (Kruijssen 2014; Adamo et al. 2015) (Portegies Zwart et al. 2010; see also Larsen 2009)
Young stellar cluster summary Stellar clusters form wherever star formation is ongoing Between 1-50% of star formation happens in clusters Fraction of stars forming in clusters increases with star formation rate surface density ICMF power law index α = 2, but requires a high-mass truncation to mass function that increases with gas pressure
What can YMCs tell us about GC formation? GC-like clusters form where gas pressure and SFR are high Both peaked near GC formation redshift (z 2) (Madau & Dickinson 2014) Are GCs the relics of regular cluster formation at high redshift?
GC population properties to explain Universal mass function turnover mass 2.2 10 5 M, no environmental dependence (Fall & Zhang 2001; Jordán et al. 2007) YMCs GCs blue GCs red GCs Bi(or multi-) modal colour/metallicity distributions (Peng et al. 2006; Usher et al. 2012)
GC population properties to explain Specific frequencies (S N = N GC 10 0.4(M V +15) ); different relations for blue and red GCs (Peng et al. 2008) Age-metallicity trends (Forbes & Bridges 2010)
GC population properties to explain GC system mass DM halo mass relation: M GC,tot /M DM 0.003% (Spitler & Forbes 2009; Durrell et al. 2014)
Previous work on GC properties Can reproduce GC mass function (Fall & Zhang 2001; Prieto & Gnedin 2008) but only with environmentally independent disruption; inconsistent with N-body simulations (Vesperini 2001; Baumgardt & Makino 2003) Hierarchical galaxy formation can produce metallicity bi/multi-modality (Tonini 2013; Li & Gnedin 2014) Models can reproduce GC spatial distributions (Prieto & Gnedin 2008; Li & Gnedin 2014)... and age-metallicity trends (Li & Gnedin 2014) but do so using incomplete physical models All studies neglect at least one of the following: Cluster formation (truncated formation redshifts; assume present day specific frequencies; GCs initialised in the halo) Cluster evolution (environmentally-independent evaporation only; no tidal shocks from gas; no spatially varying cluster disruption) Galaxy evolution (static potential; CFR independent of SFR)
Previous work on GC properties First steps toward incorporating all relevant physics to understand GC formation (Kruijssen 2015) First end-to-end model, analytic model; assumes no new physics: Star clusters form in normal disk and dwarf galaxies at high redshift (according to young stellar cluster relations) Undergo rapid disruption phase (disruption by ISM) Galaxy mergers redistribute surviving clusters Final slow evaporation phase in halo Reproduces key GC properties, including: near constant mass function turnover mass specific frequency variation with galaxy mass/metallicity GC system mass DM halo mass relation
Previous work on GC properties First steps toward incorporating all relevant physics to understand GC formation (Kruijssen 2015) First end-to-end model, analytic model; assumes no new physics: Star clusters form in normal disk and dwarf galaxies at high redshift (according to young stellar cluster relations) Undergo rapid disruption phase (disruption by ISM) Galaxy mergers redistribute surviving clusters Final slow evaporation phase in halo Conclusion: GCs are the relics of regular star formation in normal high-redshift galaxies An analytic model with limited spatially varying formation/evolution Can the results be verified and extended with more complete models?
Modelling star clusters over a Hubble time Requires: Galaxy formation (including gas): cosmological, hydrodynamical simulations Cluster formation: scaling relations of young star clusters Cluster evolution (i.e. mass loss and migration)
Modelling star clusters over a Hubble time Requires: Galaxy formation (including gas): cosmological, hydrodynamical simulations Cluster formation: scaling relations of young star clusters Cluster evolution (i.e. mass loss and migration) The E-MOSAICS project: MOdelling Star cluster system Assembly In Cosmological Simulations in the context of EAGLE (Pfeffer, Kruijssen, Crain, Bastian, Schaye, in prep.) Couple Kruijssen+11,12 star cluster formation and evolution model to the EAGLE cosmological, hydrodynamical galaxy formation simulations
The EAGLE project: Evolution and Assembly of GaLaxies and their Environments Cosmological, hydrodynamical galaxy formation simulations Largest run: 6.8 billion particles, 100 Mpc box, containing >10,000 galaxies Milky Way or bigger. Resolution 10 6 M, 0.7 kpc (Schaye et al. 2015; Crain et al. 2015)
The EAGLE project: Evolution and Assembly of GaLaxies and their Environments Model includes: Cooling and heating of the gas due to the presence of stars and other emission Formation of stars in cold and dense regions Evolution and ageing of these stars Distribution of the energy and metals (elements heavier than hydrogen) generated by the stars into the surrounding gas Explosion of supernovae with injection of their energy in the surrounding gas Formation of supermassive black holes Accretion of gas onto the black holes Ejection of energy due to this accretion process Cannot determine feedback from first principles so calibrate directly to observed galaxy properties (galaxy mass function, M,gal -M DM, M,gal -M bh, galaxy sizes)
The EAGLE project: Evolution and Assembly of GaLaxies and their Environments Schaye et al. 2015 Unique in being the only hydrodynamical simulations to reproduce the observed properties of the evolving galaxy population, in particular their stellar masses and sizes Reproduces (non-exhaustive): Specific star formation rates and passive galaxy fractions Tully-Fisher relation Mass-metallicity relations X-ray observations of the intracluster medium Column density distributions of intergalactic metals
Model for cluster formation and evolution Kruijssen+11,12 semi-analytic model: Subgrid model for the formation and evolution of the entire cluster population Cluster formation from YMC based theory using the local gas properties in the simulation ICMF power-law with index α = 2 exponential truncation mass M c(σ, ρ) cluster formation efficiency Γ(P) Cluster mass-loss model calibrated to direct N-body simulations of clusters in a tidal field (Baumgardt & Makino 2003) Disruption by tidal shocks and evaporation using the evolving local tidal field of each cluster particle Reproduces age distributions of young clusters in quiescent spiral galaxies (Adamo & Bastian 2015)
Co-formation of galaxies and GCs: the E-MOSAICS project First simulations to self-consistently model the formation and evolution of a Milky Way-type galaxy and its star cluster population
E-MOSAICS: preliminary results Cosmic cluster formation efficiency Highest peaks during galaxy mergers Break in mass function at 10 5 M
Summary Formation of GC populations has remained a major unsolved problem in astrophysics GC-like clusters form where gas pressure and SFR are high GCs are likely to be the result of regular star cluster formation at high redshift