E-MOSAICS: tracing galaxy formation and assembly with globular clusters Joel Pfeffer (LJMU) Diederik Kruijssen (ZAH), Rob Crain (LJMU), Nate Bastian (LJMU) Marta Reina-Campos (ZAH), Meghan Hughes (LJMU) Pfeffer et al. 2018 Kruijssen, Pfeffer+ 2018a, MNRAS, subm. Kruijssen, Pfeffer+ 2018b, arXiv:1806.05680 Thob, Crain, Pfeffer+, in prep.
Using globular clusters (GCs) to trace galaxy formation? Globular clusters are powerful probes of galaxy formation (e.g. review by Brodie & Strader 06) Can observe GCs to significantly larger distances than individual stars, and (potentially) obtain metallicities, ages, kinematics
Using globular clusters (GCs) to trace galaxy formation? Globular clusters are powerful probes of galaxy formation (e.g. review by Brodie & Strader 06) Can observe GCs to significantly larger distances than individual stars, and (potentially) obtain metallicities, ages, kinematics But we require a complete model for galaxy and GC formation. . .
Towards a complete model of GC formation We want to model full star cluster populations in populations of galaxies Currently not possible to simultaneously model the small scales of star/star cluster formation ( < pc) and large scales of galaxy formation ( ∼ Mpc) in hydrodynamical simulations
Towards a complete model of GC formation We want to model full star cluster populations in populations of galaxies Currently not possible to simultaneously model the small scales of star/star cluster formation ( < pc) and large scales of galaxy formation ( ∼ Mpc) in hydrodynamical simulations Requires: • A model for cluster formation (observations of young star clusters) • Cluster evolution and disruption ( N -body simulations) • Galaxy formation including baryons
Towards a complete model of GC formation We want to model full star cluster populations in populations of galaxies Currently not possible to simultaneously model the small scales of star/star cluster formation ( < pc) and large scales of galaxy formation ( ∼ Mpc) in hydrodynamical simulations Requires: • A model for cluster formation (observations of young star clusters) • Cluster evolution and disruption ( N -body simulations) • Galaxy formation including baryons Enter E-MOSAICS. . .
The E-MOSAICS project: co-formation of galaxies and GCs MO delling S tar cluster population A ssembly I n C osmological S imulations in the context of E AGLE (Pfeffer+ 18; Kruijssen+ 18) Couple sub-grid cluster model (MOSAICS) to EAGLE galaxy formation model (Schaye+ 15; Crain+ 15) Using EAGLE Recal (high-res) model (baryonic particle masses ∼ 2 × 10 5 M ⊙ ) 25 cosmological zoom-ins of Milky Way-mass galaxies (over 200 simulations in total including subgrid model testing) Near future: galaxy groups zooms ( M vir ∼ 10 13 M ⊙ ) and 34 Mpc periodic volume currently running. . . With thanks to the Virgo Consortium for DiRAC supercomputing time
MOSAICS: sub-grid model for cluster formation and evolution Kruijssen+ 11; Pfeffer+ 18: • On-the-fly modelling • Each star particle hosts its own sub-grid cluster population ⇒ Form cluster population with local bound cluster formation efficiency (CFE) at each new star particle during simulation • Schechter initial cluster mass function (power law slope − 2, with truncation M c , ∗ ), consistent with observations of YSCs • Cluster formation depends on local (gas/dynamical) properties in simulation. Completely described by CFE and M c , ∗ • Cluster mass-loss by stellar evolution, tidal shocks and evaporation using the evolving local tidal field of each ‘cluster particle’ • Dynamical friction in post-processing
The E-MOSAICS project: co-formation of galaxies and GCs First self-consistent simulations of the formation and evolution of Milky Way-type galaxies and their GC populations over full cosmic history (Pfeffer+ 18; Kruijssen+ 18)
Two main goals of E-MOSAICS Are GCs just evolved versions of young clusters? Yes Can we use GCs to trace the formation and assembly of galaxies?
Two main goals of E-MOSAICS Are GCs just evolved versions of young clusters? Yes Can we use GCs to trace the formation and assembly of galaxies?
Two main goals of E-MOSAICS Are GCs just evolved versions of young clusters? Yes Can we use GCs to trace the formation and assembly of galaxies?
E-MOSAICS: GCs are good tracers of galaxy formation Formation redshift Formation redshift 0.0 0.1 0.2 0.4 0.6 1.0 2.0 3.0 6.0 0.0 0.1 0.2 0.4 0.6 1.0 2.0 3.0 6.0 0 . 5 0 . 0 Metallicity [Fe/H] − 0 . 5 10 10 − 1 . 0 − 1 . 5 Young star clusters are the − 2 . 0 peaks of star formation in MW09 MW14 − 2 . 5 the hierarchical ISM (see 0 . 5 10 9 Galaxy stellar mass M ∗ [M ⊙ ] 0 . 0 Longmore+ 14 for a review) Metallicity [Fe/H] − 0 . 5 − 1 . 0 − 1 . 5 If GCs form like young 10 8 − 2 . 0 clusters, then GCs trace the MW15 MW18 − 2 . 5 0 . 5 enrichment history of their 0 . 0 host galaxy Metallicity [Fe/H] 10 7 − 0 . 5 − 1 . 0 − 1 . 5 − 2 . 0 MW19 MW23 − 2 . 5 10 6 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14 Age [Gyr] Age [Gyr] (Kruijssen, Pfeffer+ 2018, subm.)
E-MOSAICS: galaxy formation from GC age-metallicity relations Correlate 12 GC age-Z metrics and N GC with 30 quantities describing galaxy formation E.g. M vir , V max , c NFW , formation and assembly timescales, merger histories ⇒ Obtain 20 highly significant correlations ( p eff = p / N corr , Holm 79) (Kruijssen, Pfeffer+ 2018, subm.)
E-MOSAICS: galaxy formation from GC age-metallicity relations Correlate 12 GC age-Z metrics and N GC with 30 quantities describing galaxy formation E.g. M vir , V max , c NFW , formation and assembly timescales, merger histories ⇒ Obtain 20 highly significant correlations ( p eff = p / N corr , Holm 79) (Kruijssen, Pfeffer+ 2018, subm.)
E-MOSAICS: application to the Milky Way (Kruijssen, Pfeffer+ 2018, MNRAS, arXiv:1806.05680)
E-MOSAICS: application to the Milky Way MW had ∼ 15 mergers with galaxies M ∗ > 5 × 10 6 M ⊙ The MW assembled early for its halo mass: z a ≈ 1 . 2 (sim. mean ≈ 0 . 8) (See also Mackereth+ 18, based on stellar [ α/ Fe ]-[ Fe / H ] bimodality) (Kruijssen, Pfeffer+ 2018, MNRAS, arXiv:1806.05680)
E-MOSAICS: age-metallicity-mass relation for galaxies Can we see galaxy accretion events in the GC age-metallicity relations? (e.g. Forbes & Bridges 10; Leaman+ 13) 0.5 0.0 Galaxy enrichment history 0.5 depends on galaxy mass ⇒ higher mass galaxies enrich [Fe/H] 1.0 faster 1.5 7.5 < log 10 ( M * /M ) < 8 (Median galaxy enrichment 8 < log 10 ( M * /M ) < 8.5 8.5 < log 10 ( M * /M ) < 9 histories from EAGLE Recal) 2.0 9 < log 10 ( M * /M ) < 9.5 9.5 < log 10 ( M * /M ) < 10 10 < log 10 ( M * /M ) < 10.5 2.5 0 2 4 6 8 10 12 14 Age [Gyr]
E-MOSAICS: MW GC age-metallicity relations GC relations in age-metallicity space constrain both the galaxy mass evolution and the number of GCs per halo (at z = 0) MW accreted two galaxies M ∗ ≈ 10 9 M ⊙ with ∼ 20 GCs and one galaxy M ∗ ≈ 10 8 M ⊙ with ∼ 8 GCs (Probably more we can’t distinguish) (Kruijssen, Pfeffer+ 2018, MNRAS, arXiv:1806.05680)
E-MOSAICS: MW GC age-metallicity relations GC relations in age-metallicity space constrain both the galaxy mass evolution and the number of GCs per halo (at z = 0) MW accreted two galaxies M ∗ ≈ 10 9 M ⊙ with ∼ 20 GCs and one galaxy M ∗ ≈ 10 8 M ⊙ with ∼ 8 GCs (Probably more we can’t distinguish) (Kruijssen, Pfeffer+ 2018, MNRAS, arXiv:1806.05680)
Formation and assembly of the Milky Way from its GCs Redshift 0.0 0.1 0.2 0.4 0.6 1.0 2.0 3.0 6.0 10 11 10 10 Stellar mass [M ⊙ ] 10 9 10 8 Main progenitor Satellites 1 & 2 10 7 Satellite 3 Papovich et al. (2015) Milky Way progenitors 10 6 0.5 Metallicity of newly-formed stars [Fe/H] 0.0 -0.5 -1.0 -1.5 GCs with 10 5 < M/ M ⊙ < 10 6 . 3 -2.0 Haywood et al. (2013) Galactic disc stars Snaith et al. (2015) Galactic enrichment history -2.5 0 2 4 6 8 10 12 14 Age [Gyr] (Kruijssen, Pfeffer+ 2018, MNRAS, arXiv:1806.05680)
Formation and assembly of the Milky Way from its GCs (Kruijssen, Pfeffer+ 2018, MNRAS, arXiv:1806.05680)
Formation and assembly of the Milky Way from its GCs Canis Major = The Sausage/Gaia-Enceladus (?) Most massive galaxy ever accreted = Kraken (Kruijssen, Pfeffer+ 2018, MNRAS, arXiv:1806.05680)
E-MOSAICS: using GCs to trace shapes of MW-mass DM haloes Thob, Crain, Pfeffer+ in prep. 1 . 0 1 ExSitu vs DM in 90%GCs (Using the iterative reduced inertia tensor 0 . 9 0 . 8 method, Schneider+12) 0 . 8 20 7 21 18 M central / M FOF 10 8 11 2 23 3 15 0 . 7 12 0 . 6 9 1 0 Accreted GCs trace the DM halo 4 ǫ GCs 19 17 16 13 5 shape poorly. . . 6 0 . 6 0 . 4 14 22 24 0 . 5 0 . 2 1 . 0 1 Poor Fe / H vs DM in 90%GCs 0 . 9 0 . 0 0 . 4 0 . 0 0 . 2 0 . 4 0 . 6 0 . 8 1 . 0 0 . 8 0 . 8 ǫ DM 15 M central / M FOF 0 . 7 0 . 6 18 12 ǫ GCs 11 23 21 0 2 1 7 3 10 5 8 19 0 . 6 20 0 . 4 4 22 13 16 9 24 14 6 0 . 5 17 0 . 2 Metal-poor GCs ([ Fe / H ] < − 1) trace the shape of the DM halo very well! 0 . 0 0 . 4 0 . 0 0 . 2 0 . 4 0 . 6 0 . 8 1 . 0 ǫ DM
Recommend
More recommend
