1. Mapping the intergalactic medium in emission at low and high redshift Serena Bertone (UC Santa Cruz) Joop Schaye (Leiden Observatory) & the OWLS Team Queen’s University, Kingston April 3 rd 2009

2. Outline Introduction: the intergalactic medium: warm, warm-hot, hot… what is it? how do we detect it? New simulations: OWLS Results: X-ray & UV emission at z<1 rest-frame UV emission at z>2 what gas does emission trace? dependence on physics

3. unbiased info on matter power spectrum on largest scales reservoir of fuel for star and galaxy formation IGM metallicity constrains the cosmic SF history interplay IGM- feedback puts constraints on galaxy formation models Intergalactic medium: Why do we care? Neutrinos: 0.3% Metals: 0.03% Dark energy: 70% Free H & He: 4% Stars: 0.5% Dark matter: 25% > 90% of the baryonic mass is in diffuse gas (Persic & Salucci 1992, Fukugita et al 1998, Cen & Ostriker 1999)

4. 100 Mpc z=6 z=2z=0 Springel 2003 IGM evolution cosmic time

5. Gas temperature distribution Cen & Ostriker 1999 etc T<10 5 K10 5 K <T<10 7 KT>10 7 K Warm gas: diffused photo-ionised traced by Lyα forest Warm-hot gas: WHIM mildly overdense collisionally ionised shock heated by gravitational shocks Hot gas: dense metal enriched haloes of galaxies and clusters z=0

6. IGM mass and metals The bulk of the metal mass does not trace the bulk of the IGM mass most mass most metal mass Bertone et al. 2009a z=0.25

7. Mass and metal fractions evolution Average IGM temperature increases with time Most metals locked in stars at z=0 Average metal temperature increases with time Metal mass - Wiersma et al 2009b WHIM ICM diffuse IGM halo gas IGM mass - Dave’ et al. 2001

8. T<10 6 K Rest-frame UV lines Lyα, OVI, CIV… T>10 6 K Soft X-rays metal lines OVII, OVIII, FeXVII Absorption z<1 Nicastro et al 2005 Rasmussen et al 2007 Buote et al 2009 z<1: UV Tripp et al 2007 Lehner et al 2007 Danforth & Shull 2005 z>1.5: optical Kim et al 2001 Simcoe et al 2006 z<1 Fang et al 2005 Bertone et al 2009a z<1: UV Furlanetto et al 2004 Bertone et al 2009a z>1.5: optical Weidinger et al 2004 Bertone et al 2009b Emission how can we detect the igm? ✔ ✔ ✗ ✔✗✔✗ ✗ ✔?✔?

9. Emission vs. absorption EmissionAbsorption PRO 3-D data on spatial distribution, velocity field, metallicity… PRO Easier to detect Underdense regions investigated with current optical/UV instruments CONS Hard to detect – low fluxes Requires new instruments with high sensitivity and large FoV CONS 1-D info along a LOS Requires bright background sources

10. Owls OverWhelmingly Large Simulations The OWLS Team: Joop Schaye (PI), Claudio Dalla Vecchia, Rob Wiersma, Craig Booth, Marcel Haas (Leiden) Volker Springel (MPA Garching) Luca Tornatore (Trieste) Tom Theuns (Durham) + the Virgo Consortium Many thanks to the LOFAR and SARA supercomputing facilities

11. Owls cosmological hydrodynamical simulations: Gadget 3 many runs with varying physical prescriptions/numerics run on LOFAR IBM Bluegene/L WMAP 3 cosmology largest runs: 2x512 3 particles two main sets: L=25 Mpc/h and L=100 Mpc/h boxes evolution from z>100 to z=2 or z=0

12. New physics in Owls New star formation ( Schaye & Dalla Vecchia 2008 ) : Kennicutt-Schmidt SF law implemented without free parameters New wind model (Dalla Vecchia & Schaye 2008) : winds local to the SF event hydrodynamically coupled Added chemodynamics (Wiersma et al. 2009) : 11 elements followed explicitly (H, He, C, N, O, Ne, Si, Mg, S, Ca, Fe) Chabrier IMF SN Ia & AGB feedback New cooling module (Wiersma, Schaye & Smith 2009) : cooling rates calculated element-by-element from 2D CLOUDY tables photo-ionisation by evolving UV background included

13. Physics variations in OWLS Cosmology: WMAP1 vs WMAP3 vs WMAP5 Reionisation & Helium reionisation Gas cooling: primordial abundances vs metal dependent Star formation: top heavy IMF in bursts isothermal & adiabatic EoS Schmidt law normalisation Metallicity-dependent SF thresholds Feedback: no feedback feedback intensity: mass loading, initial velocity… feedback implementation (Springel & Hernquist; Oppenheimer & Dave’) AGN feedback Chemodynamics: Chabrier vs Salpeter IMF SN Ia enrichment AGB mass transfer

14. Gas cooling rates Wiersma, Schaye & Smith 2008 collisional ionisation eq. photoionisation eq. density dependent Photo-ionisation by UV BK + collisional ionisation equilibrium Cooling rates calculated element by element for 11 species: takes into account changes in the relative abundances

15. Gas emissivity 11 elements – comparable to cooling rates Collisional ionisation + photo-ionisation by UV BK  important at low density UV lines stronger than X-ray ones UV linesX-ray lines Density  Bertone et al 2009a z=0.25

16. Emission at low redshift: Soft X-rays & UV lines

17. Emission at low redshift: X-rays 100 Mpc/h boxes 20 Mpc/h thick slices 15” angular resolution 12 emission lines Bertone et al 2009a

18. X-ray lines O VIII strongest line lines from lower ionisation states and whose emissivity peaks at lower temperatures trace moderately dense IGM: C V, C VI, N VII, O VII, O VIII and Ne IX lines from higher ionisation states trace denser, hotter gas: C VI, O VIII, Ne X, Mg XII, Si XIII, S XV and Fe XVII Fe XVII emission has different spatial distribution than other elements: later enrichment by SN Ia Bertone et al 2009a

19. What gas produces x-ray emission? X-ray emission traces warm-hot, moderately dense gas, not the bulk of the IGM mass and metals. emission-weighted particle distributions Bertone et al 2009a

20. Emission at low redshift: UV Bertone et al 2009a UV emission is a good tracer of galaxies and of mildly dense IGM, but not of IGM in very dense environments

21. UV lines C IV strongest line: traces gas in proximity of galaxies different spatial distribution: O VI and Ne VIII trace more diffuse gas than C IV no emission from the hottest gas in groups

22. What gas produces UV emission? emission-weighted particle distributions Bertone et al 2009a O VI and Ne VIII trace WHIM gas C IV traces more metal-rich, cooler gas

23. Summary: Dependencies on gas properties Correlation of emission with density and metallicity: highest emission from densest and most metal enriched particles Median temperature of highest emission corresponds to peak temperature of emissivity curve – as seen in T-n HI diagrams Median density, temperature and metallicity of gas particles vs. particle emission Bertone et al 2009a

24. Impact of physics

25. What happens when changing the physical model? Impact of physics: x-rays no feedback: no metal transport  localised emission primordial cooling rates: longer cooling times  stronger emission at high density (≈100 times) momentum driven winds: metals more spread in IGM  weaker emission (≈100 times) AGN feedback: weaker emission in dense regions Bertone et al 2009a

26. no feedback: emission localised in galaxies primordial cooling rates: stronger emission at high density AGN feedback: weaker emission changes in wind parameters: small effect Impact of physics: UV Bertone et al 2009a

27. Emission at high redshift: rest-frame UV lines Bertone et al 2009b emission at 2<z<5 25 Mpc/h simulations 2” angular resolution

28. IGM emission at z>1.5 At z>1.5 rest-frame UV lines are redshifted in to the optical band A number of upcoming optical instrument might detect IGM emission lines at 1.5<z<5: Cosmic Web Imager on Palomar (CWI, Rahman et al 2006)  this year! Keck Cosmic Web Imager (KCWI) Antarctic Cosmic Web Imager (ACWI, Moore et al 2008) MUSE on VLT (Bacon et al 2009) IFUs with large fields of view & high spatial resolution Great chance to observe the 3-D structure of the IGM for the first time!

29. Higher ionisation states

30. Lower ionisation states

31. Emission PDFs Lower ionisation states: single lines shorter wavelengths C III up to 10x stronger than C IV Higher ionisation states: doublets – easy to identify weaker than lower ion. states

32. What gas produces UV emission? emission-weighted particle distributions C and Si lines trace colder, more metal enriched gas than O, N and Ne lines Most emission comes from dense gas  more metal enriched More highly ionised lines trace warmer gas (e.g. C IV vs C III)

33. What can we observe with CWI? CWI: flux limit: 100 photon/s/cm 2 /sr angular resolution: 2” Everything in red and white might be detectable  central regions of groups and galactic haloes z=2

34. Lines detectable by CWI LineWavelength (Å) Minimum redshiftDetectable to z C III977.02.7≤ 4 C IV15481.4≤ 3 N IV7653.7× N V1230.91.9≤ 2 O IV549.35.6× O V630.04.7× O VI10322.5≤ 3 Ne VIII770.03.7× Si III12072.0≤ 3 Si IV13941.6≤ 3

35. summary Dense cool gas in the haloes of galaxies is traced by: UV lines: C IV, Si IV Low density WHIM gas in filaments is traced by: UV lines: O VI, Ne VIII soft X-ray lines from hydrogen-like atoms: C V, O VII, Ne IX soft X-ray lines from elements with low atomic numbers: OVIII Dense hot gas in clusters and groups is traced by: soft X-ray lines from fully ionised atoms: C VI, OVIII soft X-ray lines from elements with high atomic numbers: Mg XII, Fe XVII Detection of WHIM emission by future telescopes: challenging in low density regions very likely in groups and cluster outskirts CWI very likely to detect metal line emission at high z for the first time

37. Convergence tests Angular resolution: needed for detecting the spatial distribution of the emitting gas Number of particles: need high resolution for physical processes (star formation, feedback, cooling etc) to converge

38. Emission per unit volume Bertone et al 2009b Most energy emitted by lower ionisation lines: C III, O IV, O V O VI strongest line at higher ionisation Lot of energy emitted by Si lines, despite low emissivity Ne and N emission per unit volume very weak

40. strongest X-ray emission from dense regions weak emission from the low density, metal poor gas Emission vs density Density cut maps