1. Alexandra Pope (UMass Amherst) JWST Workshop – STScI Baltimore June 8, 2011 Mid-Infrared Observation of High Redshift Galaxy Evolution

2. Most star formation activity occurs behind dust Chary & Pope 2011 Bouwens et al. 2009 Lagache et al. 2005

3. Most stellar mass and AGN growth occurs at z~1-3 sub-mm QSOs Squares: optical/near-IR Circles: far-IR/sub-mm Wall et al. 2008 Dickinson et al. 2003

4. Le Floc’h et al. 2005 ULIRGs LIRGs Normal Total Most of the z~1-3 stellar mass growth occurs in LIRGs (and ULIRGs)

5. Murphy et al. 2011

6. Ultra-luminous infrared galaxies (ULIRGs) “A rose by any other name would smell as sweet ” Local Complete samples of galaxies down to a given L IR Not biased by AGN, dust temperature, detection limit, etc. Mostly major mergers z ~ 2

7. Ultra-luminous infrared galaxies (ULIRGs) “A rose by any other name would smell as sweet ” Local Complete samples of galaxies down to a given L IR Not biased by AGN, dust temperature, detection limit, etc. Mostly major mergers z ~ 2

8. Submillimeter Galaxies (SMGs) high redshift z~2 (Chapman et al. 2005, Pope et al. 2005) massive M gas ~3·10 10 M  M  ~10 11 M  (Greve et al. 2005, Borys et al. 2005) high IR luminosities => SFR~1000M  yr -1 … assuming no AGN! (Lilly et al. 1999, Chapman et al. 2005, Pope et al. 2006) sub-mm Borys et al. 2005 Pope et al. 2005 Chapman et al. 2005

9. Spectral energy distribution (SED) of high redshift submillimeter galaxies (SMGs) Rest Wavelength (  m) Luminosity density (WHz -1 ) HST VLA SCUBA MIPS IRAC Herschel IRS Circa 2011: Well sampled SED for ULIRGs z~1-3 Spectroscopy can provide a probe of the underlying radiation field Mid-IR spectroscopy is sensitive to dust which is dominating the SFRD

10. Outstanding issues: Projects for JWST Resolve the 24 μ m background – detect warm dust in faint LBGs which contribute significantly to the IR background; SMGs out to z~7 Detect PAH emission in lower luminosity galaxies (<10 12 L sun ) and/or out to z~7 - trace SF and ISM Resolve PAH emission at z~1-3 (?) Probe lower luminosity, obscured AGN in (U)LIRGs - High ionization lines Rigby talk Trace shocks and turbulence in galaxies - Warm H 2 emission Guillard talk

11. Resolve the 24 μ m background – detect warm dust in faint LBGs which contribute significantly to the IR background; SMGs out to z~7 Detect PAH emission in lower luminosity galaxies (<10 12 L sun ) and/or out to z~7 - trace SF and ISM Resolve PAH emission at z~1-3 (?) Probe lower luminosity, obscured AGN in (U)LIRGs - High ionization lines Trace shocks and turbulence in galaxies - Warm H 2 emission Outstanding issues: Projects for JWST

12. Deep Spitzer MIPS 24 μ m surveys Spitzer GOODS Legacy Survey Figure from M. Dickinson

13. Deep Spitzer MIPS 24 μ m surveys Secure detections * account for 70% of the 24 μ m background (Papovich+04, Chary+04) e.g. Counterparts to submillimeter galaxies (SMGs) out to z~4 (Pope+06) e.g. Detect warm dust in 2/3 of LBGs (Reddy+08) * Reach the confusion limit at ~60 μ Jy (5 σ, Dole+2004 ) Spitzer GOODS Legacy Survey

14. Deep Spitzer MIPS 24 μ m surveys Secure detections * account for 70% of the 24 μ m background (Papovich+04, Chary+04) e.g. Counterparts to submillimeter galaxies (SMGs) out to z~4 (Pope+06) e.g. Detect warm dust in 2/3 * of LBGs (Reddy+08) * Reach the confusion limit at ~60 μ Jy (5 σ, Dole+2004 ) Pope et al. 2006, 2008

15. Many star forming galaxies still missing from deep 24 μ m surveys Reddy et al. 2010 Bouwens et al. 2009

16. Deep JWST MIRI 24 μ m surveys Resolve >99% of the 24 μ m background (Dole+2004) e.g. Counterparts to all submillimeter galaxies (SMGs) out to z~7 e.g. Detect warm dust in 100% of LBGs e.g. A few surprises? Simulated JWST image 7 times the spatial resolution of Spitzer

17. Outstanding issues: Projects for JWST Resolve the 24 μ m background – detect warm dust in faint LBGs which contribute significantly to the IR background; SMGs out to z~7 Detect PAH emission in lower luminosity galaxies (<10 12 L sun ) and/or out to z~7 - trace SF and ISM Resolve PAH emission at z~1-3 (?) Probe lower luminosity, obscured AGN in (U)LIRGs - High ionization lines Trace shocks and turbulence in galaxies - Warm H 2 emission

18. (ISO) Mid-infrared Spectroscopy Local ULIRGs Lutz et al. 1998 Genzel et al. 1998

19. Spitzer IRS spectroscopy Detailed studies of local galaxies Smith et al. 2007

20. 1.Starburst : Polycyclic aromatic hydrocarbons (PAH) emission lines + extinction Main lines at 6.2, 7.7, 8.6 and 11.3  m 2. AGN: power-law + extinction Spitzer Mid-IR Spectroscopy Extending out to high redshift The mid-IR spectrum can be decomposed into two main components: Weedman et al. 2006

21. z~2 Submillimeter Galaxies (SMGs) high IR luminosities => SFR~1000M  yr -1 … assuming no AGN! Local ULIRGs show increasing AGN contribution with L IR => Does this mean SMGs are AGN dominated ? Log [L IR (L  )] SB fraction SMGs Tran et al. 2001

22. Starburst: Polycyclic aromatic hydrocarbons (PAH) + extinction AGN: power-law + extinction Spitzer Mid-IR Spectroscopy of SMGs SMG composite spectrum Mid-IR SED of SMGs is starburst dominated : small contribution from AGN (<30% at these wavelengths) Scaled up M82 – not like local ULIRGs (e.g. Arp220) Pope et al. 2008 see also Valiante et al. 2007, Menendez-Delmestre et al. 2009

23. SMGs (Pope et al. 2008) Local BGS ULIRGs (Armus et al. 2007) Local SB (Brandl et al. 2006) Pope et al. 2008 SMGs are scaled up local starbursts

24. Spitzer-selected ULIRGs Fadda et al. 2010 Spitzer-selected (U)LIRGs at z=1 and z=2 also show higher PAH EW than local ULIRGS Stronger PAH emission at high redshift (for a given L IR )

25. IRS-identified Compton-thick AGN Pope et al. 2008 Strong power-law continuum (f  -2 ) No X-ray detection!  Compton-thick AGN: N H >> 10 24 cm -2 See Alexander et al. 2008 for more discussion on IRS identified Compton-thick AGN

26. What about lower luminosity galaxies? Siana et al. 2008, 2009; see also Rigby et al. 2008, Gonzalez et al. 2010

27. JWST/MIRI spectra of z~2 LIRGs Spitzer IRS took spectra of ~200 galaxies at z>1 - the majority of these are ULIRGs, with a few LIRG – learnt a lot of AGN vs SF in the most luminous galaxies JWST/MIRI will be able to take spectra of all LIRGs and at much higher spectral resolution (R~3000!) Armus et al. 2007 Spitzer high resolution (R~600) only possible for local galaxies

28. Getting more info out of a PAH spectrum Relative PAH strengths tell you about the size distribution and ionized state of the dust grains; compared to large grains you can constrain the strength of the radiation field Draine & Li 2007 Smith et al.2007

29. Getting more info out of a PAH spectrum Draine & Li 2007 Pope et al. 2008

30. JWST/MIRI spectra of z>4 galaxies? Spitzer/IRS saw PAHs out to z~4 JWST/MIRI will see the 6.2 μ m PAH out to z~3.5 and the 3.3 μ m PAH out to z~7 – track down the earliest dust and learn how it formed 24 hr Spitzer IRS spectrum of the z=4 SMG GN20 Riechers et al. in preparation

31. 3.3 μ m PAH feature in local galaxies EW provides a cleaner AGN indicator Imanishi et al. 2010

32. 3.3 μ m PAH at high redshift Sajina et al. 2009

33. What if SMGs are really powered by AGN? In order to have most of the L IR from the AGN and have the PAHs survive, you need to have an unusual geometry, e.g. non PAH dust shielding more PAH dust from AGN radiation or just really extended distribution of small grains Can we resolve this??

34. What if SMGs are really powered by AGN? SMGs have sizes of : ~0.6 arcsec ~5kpc ( Engel+10, Younger+08 ) JWST/MIRI has a resolution of 0.3 arcsec at 10 μ m – we can try to resolve the 3.3 μ m PAH emission in galaxies at z<2 Compare to distribution of large dust grains as traced by ALMA

35. Warm molecular Hydrogen Traces shocks and turbulence Gonzalez et al. 2010

36. Synergy with Herschel, ALMA and other (sub)mm facilities We need to plan JWST program that will complement other capabilities and facilities circa 2017/2018: Herschel will have provided a statistic samples of high redshift ULIRGs ALMA will be underway “detecting molecular gas in Milky Way like galaxies out to z~3” – but on smaller targeted samples Large single dish (sub)mm telescopes (LMT, CCAT) can provide statistical samples of LIRGs – only probing the cold dust

37. Herschel Space Observatory Deep high(er) resolution far-IR imaging at: 70, 100, 160, 250, 350, 500μm  Accurate L IR and SFR for individual galaxies  Dust properties: T dust, β, M dust GOODS Herschel Key Program PI David Elbaz

38. GOODS Extragalactic Mid-IR Spectral Library 150 sources in GOODS-N and ECDFS observed with Spitzer/IRS All data reduced and lines measured in a uniform way Spectral decomposition to get AGN fraction, L PAH, L 8, etc. AGN < 50% AGN > 50%

39. SED fitting with Spitzer/IRS + Herschel data AGN dominated (mid-IR)SF dominated (mid-IR) Warm dust (100K) Cool dust (~30K)

40. Composite SEDs Kirkpatrick et al. in prep

41. Composite SEDs

42. Influence of the AGN on IR colors Hatziminaoglou al. 2010 Kirkpatrick et al. in prep

43. Valtchanov et al. 2011 Synergy with ALMA: Constraining ISM conditions

44. Valtchanov et al. 2011 Synergy with ALMA: Constraining ISM conditions

45. Summary We have learnt a lot from Spitzer – this puts us in a great position to plan innovative projects for JWST Several key areas where still have a lot to learn about dust in high redshift galaxies: Directly detect dust in the galaxies which dominate the SFRD Understand the detailed composition of dust at high redshift which constrains the production mechanisms Relative heating of dust from starbursts and AGN activity JWST together with ALMA will be a powerful duo for attacking these problems