Day 3. Cosmic chemical evolution.! Chiaki Kobayashi! (Univ. of Hertfordshire, UK)!

Transcription

1 Day 3. Cosmic chemical evolution! Chiaki Kobayashi! (Univ. of Hertfordshire, UK)!

2 Contents Day1: Nucleosynthesis of massive stars & supernovae! Day2: Galactic Chemical Evolution (GCE)! Day3: Cosmic chemical evolution! Motivation! Basic equations of Chemodynamical Simulations! Test: Isolated disk simulation! Active galactic nuclei (AGN) feedback! Cosmological Simulations! Metallicity of galaxies! Cosmic evolution! Time evolution of Mass metallicity relation (MZR), Blackhole-galaxy mass (M BH -σ) relation! Metallicity radial gradients!

3 History of the Universe background radiation satellite Planck! 4.9% Atoms 26% Dark Matter 69% Dark Energy 13.8!

4 Galaxy Merger

5 Galactic Winds

6 3 types of galaxy models One-zone Semi-analytic! Chemo-dynamical simulations (monolithic) models (hierarchical) (Burkert Inhomogeneous & Hensler chemical 87, Katz 92, (Tinsley Instantaneous 80, Timmes+ mixing models enrichment Navarro & White 93, Steinmetz & 95, approximation! Pagel 97, (Kauffmann, Mass assembly White Müller Internal 94, structures Mihos & Hernquist - 94, Matteucci SFR/inflow/outflow 01, Prantzos + history 93, Cole+ based 94, on kinematics Kawata & of Gibson stars/gas, 03, CK 04, + with 93, analytic Chiappini+ formula! 97, Somerville λcdm scenario! & spatial CK & distribution Nakasato 11, of CK+ Average 00 )! evolution of a Primack Global properties 99, De elements Scannapieco+ 11, GASOLINE, galaxy (or a shell of Lucia+ of galaxies 04,! in a RAMSES, AREPO )! galaxy)! Nagashima+ large scale 05 ) Other types of models! Stochastic models (Argast+02; Ishimaru & Prantzos; Cescutti+; Wehmeyer+)! Chemodynamical models without hydro (e.g., Minchev & Chiappini)!

7 Basic equations of Chemodynamical Simulations CK 2004! CK, Springel, White 2007! Taylor & CK 2014

8 N-body methods Gravity for DM and stars! direct summation O(N 2 )! GRAPE, GPU!! j i d 2 x m i x j x i i = Gm dt 2 i m j 3 x j x i Tree method O(N logn)!!! distant particles are bundled up! Particle-mesh method O(N logn) 2 φ = 4πGρ k 2 φ k = 4πGρ k φ(r) k

9 SPH method Smoothed Particle Hydrodynamics for gas (Gingold & Monaghan 1977; Lucy 1977; Monaghan 1992 for a review)! equations to integrate:! Dρ Dt + ρ v = 0 Dv Dt = 1 P Φ ρ Du Dt = P Dρ (κ T) + ρ 2 Dt ρ 2 Φ = 4πGρ P = (γ 1)ρu ρ i = m j W (r i r j ;h) Dv i Dt = Du i Dt = m j m j + Γ Λ ρ $ P i 2 ρ + P j 2 i ρ + Π ' & ij ) iw (r i r j ;h) Φ % j ( ( ) i $ P i 2 ρ + 1 i 2 Π ' & ij )( v i v j ) i W (r i r j ;h)+ Γ Λ % ( ρ SPH formulation (Navarro & White 1993)! spline kernel W! smoothing length h: variable both in space & time! integration: Leap-frog method! individual timestep! SPH! f (r) = f ( r!)w(r r!;h)d r! f (r) = f ( r!) W(r r!;h)d r! m f (r) = j ρ(r j ) f (r )W(r r ;h) j j f (r) = m j ρ(rj ) f (r j ) W(r r j ;h)

10 Radiative Cooling Calculated by MAPPINGS III (Sutherland & Dopita 1993).! Molecule cooling (T<10 4 K) is not included.! UV background heating is subtracted (Kats, Weinberg & Hernquist 96).!

11 Katz (1992)! converging flow! rapid cooling! Jeans unstable! Star Formation Density criteria (Stinson+06; Governato+10, Nature)!!ρ > ρ crit! Star formation timescale, c= ! Stochastic treatment for the timestep Δt!! Form stars if random number [0,1] is smaller than!!!!,!

12 Chemical Enrichment 1/2 Ejection of mass at t! time! Ejection of metals at t! Kobayashi et al. 06, SNII/HN, Z-dependent Nomoto et al. 97, W7

13 Chemical Enrichment 2/2 SW rate! SN II/HN rate! time!! SN Ia rate (SD scenario)! metallicity effect! primary WD secondary star IMFs of single and binary (secondary) stars

14 Supernova Feedback 1/2 Energy by stellar winds for given mass!! Energy per core-collapse supernova! 1 e e, II = m 2,u p e,m m φ(m)dm m t % p e,m = 1 (m 2, < m 20M sun ) & (1 ε HN )+[10,10, 20,30]ε HN (20, 25, 30, 40M sun ) 1051 (erg) ' ε HN = [0.5, 0.5, 0.4, 0.01, 0.01] for Z = [0, 0.001, 0.004, 0.02, 0.05] Energy per Type Ia supernova!

15 Supernova Feedback Total energy, mass, metal ejection from a star particle j at time t to a neighbor gas particle i! E e, j (t, Z j ) = m * 0 [e e, SW (Z j )R SW (t)+ e e, II (t, Z j )+ e e, Ia R Ia (t, Z j )] = E m, j (t, Z j ) = m * 0 [e m, SW (t, Z j )+ e m, II (t, Z j )+ e m, Ia (t, Z j )] = E Z, j (t, Z j ) = m * 0 [e Z, SW (t, Z j )+ e Z, II (t, Z j )+ e Z, Ia (t, Z j )] =! Internal energy (and mass, metal) that gas particle i get from star particle j! du i dt = j 100% thermal! t t dt j dt N FB i=1 N FB i=1 E m,i E Z,i N FB i=1 E e,i E e, j (t, Z j )W (r i r j ;h j ) N FB k=1 W (r k r j ;h j ) (t, Z j )W (r i r j ;hj) (t, Z j )W (r i r j ;hj) (t, Z j )W (r i r j ;hj)

16 Other feedback methods artificial cooling off during heating (Governato+ 10)! kinetic feedback (Navarro & White 93, Springel & Hernquist 03)! v wind = 2ε E kin SN m kinetic feedback (Navarro & White 93, Springel & Hernquist 03)! m v wind = 484 km/s for p (1 η exp[ Δt ]),η = 2! m g,0 / N * t sf momentum driven-winds (Dave & Oppenheimer 06)!! v wind = 3σ f L 1 + 2σ, η = σ 0 σ = 2, σ = 300 km/s, f = 2 or 0 L stochastic treatment of feedback (Dalla Vecchia & Schaye 12)! Δε (~ erg/s) is given if random number [0,1] is smaller than!, f th =1,N ngb =48!

17 Density-temperature diagram CK, Springel, & White 2007!

18 Isolated Disk Fixed NFW halo M tot =10 10 /h M!! CK, Springel, & White 2007! 10% gas, N= (m gas = /h Mu), λ=0.1! Face-on! Edge-on Colors: Gas Temperature, White: Stars!

19 Wind Fraction Wind is defined as r>2r 200! No FB No FB FB FB FB + Z-Cooling FB High + Z-Cooling resolution High resolution Ejected! Metal Mass!! M z,w / M z,tot! 44% (SN)! 3% (SN)! 60% (HN)! 25% (HN)!! 2/3?! (Renzini 03)! ~50% gas, ~25% metals are ejected by HNe.! Resolution independent.!

20 Other Simulation Codes Cosmological! Springel 00 (Gadget - SPH), Di Matteo (AGN), Scannapieco (Ia), Schaye! Frenk, Okamoto, (Gadget)! Dave, Oppenheimer (Gadget)! Mosconi, Tissera, et al. 01 (AP3MSPH, Ia)! Galactic! Steinmetz & Muller 94 (SPH)! Raiteri, Villata & Navarro 96 (Ia)! Carraro, Lia & Chiosi 98 (Ia)! Hensler 90, Recchi! Burkert, Naab 99 (N-body)! Hernquist, Hopkins 05 (AGN)! Mori 97! Teyssier 02, Devriendt (RAMSES AMR)! Nakasato 03! Wadsley 04 (GASOLINE SPH), Gibson, Kawata 04 (Ia), Brook! Governato, Brook! Tornatore et al. 04 (Ia)! Martinez-Serrano 08 (SPH)! Elmegreen! Springel 10 (AREPO)!!!

21 AGN feedback

22 Active Galactic Nuclei (AGN) 1950s: radio sources discovered, optical counterparts?! 1960: 3C48, blue star with broad emission lines! 1963: 3C273, redshift z=0.158 distant object! z λ obs λ emit 1! very luminous, L sun,~ erg/s, energyproduction mechanism?!

23 The Unified Model of AGN Seyfert 1 with broad & narrow emission lines Seyfert 2 with! narrow emission lines

24 AGN Feedback?! AGN not only suppress SF but also enhance SF!! Wagner & Bicknell 2011; and many!

25 Co-evolution of SMBH & galaxies Quasar space density! Schmidt+ 95! Cosmic SFR! Madau+ 96! BH mass-bulge mass relation! Magorrian+ 98!

26 Stellar + AGN winds M tot ~0.5-1x10 12 M!! M * ~5x10 10 M!?! M BH ~5x10 7 M!! SFR~0?! M tot ~10 10 M!! SFR~10M! /yr! M halo ~ x10 12 M!! M bulge ~10 9 M!?! M BH ~10 6 M!! SFR(ring)~0.05M! /yr! ESO optical/submm/xray!

27 AGN feedback in Semi-Analytic Models Croton et al. 2006, Semi-analytic Model! quasar mode: BH growth by mergers! radio mode: heating by hot gas accretion on BH!!!!!!!!!!! No such modes in hydro-simulations (Springel, Di Matteo, Hernquist 05; Booth & Schaye 09; Taylor & Kobayashi 14)!

28 Standard AGN model in hydro-simulations Co-evolution of BH-galaxy is assumed.! BH Formation (Booth & Schaye 09; also Springel, Di Matteo+ 05, Sijacki+07)! Regularly run a friends-of-friends group finder on all of the dark matter particles during the simulation, get DM halo mass.! If the halo is >M halo,min =100m DM (~10 10 M! ) and does not already contain a BH, then its most gravitationally bound baryonic particle is converted into a collisionless BH particle with M seed =0.001m g (~10 5 M! ).! Growth! Feedback!

29 The first BHs? The first chemical enrichment is driven by ~10-50M! stars (faint SNe), leaving ~10M! BHs.! No signature of ~ M! stars (PISNe) observed in any chemical abundances.! If the accretion is suppressed, <100M! stars form (Hosokawa+ 11). Otherwise, >300M! stars may form (Ohkubo+ 09), which collapse to M! BHs.! Direct collapse of gas can form ~10 5 M! BHs (e.g., Madau & Rees 01), but rare.! What is the role of these first BHs in galaxy formation?! Can they grow into super-massive BHs at present?! Are they enough for the down-sizing?!

30 New AGN model BH Formation - not merging products, but originated from primordial SF! then M seed (~1000M! )! Growth - accretion & mergers (same as in previous works; Springel, Di Matteo+ 05, Booth & Schaye 09)!! Feedback thermal!

31 Cosmological Simulations CK, Springel, White 2007! Taylor & CK 2014, 2015abcd

32 Nbody+SPH+ UV background radiation! (Haardt & Madau 1996)! BH,NS,WD! E Fe O Fe E O SNIa! SD(CK+98,09)! E Cooling(Z)! O(Sutherland & Dopita 93)! Fe O Fe E O Fe E SNII/HN (CK+06)! O Fe E E Star Formation! v<0, t cool <t dyn, t dyn <t sound! t sf =t dyn /c, c=0.1, Kroupa IMF! E Stellar Wind! E Feedback! 100% thermal! to N FB ~400! E BH Formation! Z=0, ρ>ρ crit,1000m!! E BH Growth! accretion Bondi-Hoyle! merger! BH BH M acc! E

33 Cosmological Simulation Stellar Luminosity Gas Metallicity 25Mpc; N~ ; m gas ~10 7 M! ; H 0 =70, Ω m =0.28, Ω λ =0.72, Ω b =0.046, n=1, σ 8 =0.82

34 Constraining parameters from cosmic SFRs, Magorrian relation, and galaxy size-mass relation α =1, ε f =0.25, ρ=0.1, M seed =1000M!! cosmic SFR! Blackhole Mass! redshift! ~ Galaxy Mass' Taylor & CK 2014, MNRAS, 442, 2751!

35 M BH seed ~ M! better than 10, M! First Star origin Cosmic SFR! Observational Constraints Magorrian relation! 10,25/hMpc 3 run! z = 0! Taylor & CK 2014, MNRAS, 442, 2751; Taylor & CK 2015a, MNRAS, 448, 1835! Size-mass relation! obs: Bouwens+11, Karim+11, Cucciati+12, Oesch+12, Burgarella +13, Gunawardhana+13, Sobral+13! obs: Kormendy & Ho 2013!

36 Simulation Outcome Not used for parameter determination! Stellar Mass Function! Star Formation Main Sequence!

37 Other Big Simulations EAGLE w Gadget-3 (100Mpc)! Cosmic SFR?! Illustris w AREPO (106.5Mpc)! Size-Mass relation?! obs: Omand et al. 2014! Schaye et al. 14! Snyder et al. 15!

38 38! 38! Colour: Temperature; Taylor & CK 2014!

39 AGN-driven outflows Taylor & CK 2015b, MNRAS, 452, L59! AGN winds! AGN-driven winds remove the remaining gas (3% of baryon) and ~2% of total produced metals.!

40 Metallicity of galaxies

41 Star Forming Galaxies Emission line ratios [OII],Hβ, [OIII],Hα,[NII],[SII]! Photoionization models - MAPPINGS (Sutherland & Dopita 199), Cloudy (Ferland et al. 1998)! Stellar population synthesis models - Starburst99 (Leitherer et al. 1999)! Degeneracy between Z and ionization parameter q! Solar Abundance = 8.69 (Allende Prieto et al. 2001), 8.66 (Asplund et al. 2004)! Metallicity =! 12 + log (O/H)! Kewley & Ellison 08

42 Early-type Galaxies (ETG) Integrated Absorption Lines Lick indices Mg 2, Mg b, Fe5270, Fe5335, Hβ, Hγ' Population synthesis models Worthey 1994, Thomas, Maraston & Bender 2003! Broadening of velocity dispersion! Contamination of emission lines! Age-Metallicity Degeneracy! Metallicity [M/H]! Worthey 1996

43 [X/Fe]-[Fe/H] relations of ETGs Nomoto, CK, Tominaga 2013, ARAA!

44 Abundances at high-redshifts Lyman-break Galaxies QSO absorption line systems! LBG! IGM! QSO! Scale Dependence z=3! Pettini 2002!

45 The origin of MZR Gas in galaxies, identified by FOF! Wind Gas, is not in galaxy,! but has been in some galaxies! M * /M b =0.10! M g,gal /M b =0.10! M w /M b =0.20! CK+ 07!

46 The origin of MZR Metals are effectively ejected from (present) dwarfs.! The origin of the mass-metallicity relation --- Massdependent Galactic Winds.! CK+ 07!

47 Mass Metallicity Relations (MZR) Taylor & CK 2015a, MNRAS, 448, 1835! Gas-phase Oxygen Abundance,! SFR weighted! Stellar Metallicity, luminosity weighted! AGN no AGN AGN no AGN AGN does not affect MZR.!

48 [α/fe]-mass relation of ellipticals longer SF! more SNe Ia!! Taylor & CK 2015, MNRAS, 448, 1835! ~ Mass of galaxies!

49 Other models Matteucci 1994, GCE! Inverse wind?! Arrigoni et al. 10, SAM! flat IMF (x=1.1)! yields: Woosley & Weaver 95! SNIa: bimodal DTD (inconsistent with obs. in the solar neighborhood)! Calura & Menci 09, SAM! SFR-dependent IMF! yields: Woosley & Weaver 95! SNIa: Matteucci & Greggio 86! Yates et al. 13, SAM! Chabrier (03) IMF! yields: Portinari+ 98! SNIa: power-law DTD!

50 Cosmic evolution

51 Cosmic evolution Time evolution of scaling relations (Size/metallicity/BH mass galaxy mass)! JWST (>2018)!! Internal structures (2D map of gas, stars, and elements) within galaxies! IFU (Integral field unit)!

52 Time evolution of M-Z relation Gas: Steeper slope is at higher-z.! Taylor & CK 2016a, arxiv ! Stars: Normalization shifts.!

53 Time evolution of M BH -σ relation Taylor & CK 2016, ArXiv ! But M! (Wu+ 15)! Larger Volume?! Larger Seed??!

54 Time evolution of M BH -σ relation Taylor & CK 2016, ArXiv !

55 Metallicity gradients Kobayashi & Arimoto (1999) Also, Organo+ 2005, Spolaor+ 10,! Kuntschner+ 10, CALIFA, SAMI, MaNGA

56 Metallicity Gradients vs Mass Steep Kobayashi & Arimoto 1999! Spolaor, CK et al. 2010! GMOS-slit observations! GRAPE SPH simulations (CK 2004) Flat Kobayashi & Arimoto (1999) SAURON (Kuntschner et al. 2010)! Hopkins et al. 09 IFU surveys:! ATLAS 3D, CALIFA, SAMI, MaNGA!

57 Metallicity Gradients vs Mass ATLAS 3D, CALIFA, SAMI, MaNGA! Taylor & CK 2015d, submitted! Flat Spolaor (2010) s transition MW Steep Measured in MAX(2re,10kpc), V-luminosity/SFR weighted!

58 Future Galaxies consist of stars. The formation and evolutionary histories will be revealed from their chemical evolution ( g alactic archaeology).! James Webb Space Telescope (JWST)

59 Now Meantime, let s update our knowledge of stellar physics and the Milky Way, from elemental abundances and kinematics of stars with precision spectroscopy.! GAIA spacecraft