Diapositive 1

1
Mid-IR selection of Ultra-Luminous Far-IR
Galaxies Starburst and AGN tracers in z2
ULIRGs continuum CO mm emission, radio and PAH
Alain Omont (IAP, CNRS and Université Paris
6)
2
OUTLINE Mid-IR luminous Ultra-Luminous Far-IR
Galaxies

• Submillimeter galaxies high-z Ultra-Luminous
  Far-IR Galaxies
• 24µm bright z2 ULIRGs AGN vs PAHs/starburst
• Mm continuum emission. Far-IR luminosity, star
  formation rate
• Radio emission starburst/AGN
• CO mm emission. Molecular gas, structure and
  dynamics
• Stellar mass
• Sructure (HST, radio, CO) merger/outflow
• Prospects (Herschel, ALMA, JWST)
• Collaborators
• C. Lonsdale, M. Polletta, N. Fiolet, A. Beelen,
  A. Baker, F. Owen, S. Berta
• L. Yan, L. Tacconi, D. Lutz, A. Sajina , G.
  Lagache, D. Shupe

3
Reminder SMGs strongest starbursts in the
Universe Essential steps of star formation in
massive galaxies at z gt 2

• Revealed by SCUBA surveys at 850µm ( MAMBO at
  1.2mm ? AzTEC,
• LABOCA, BOLOCAM)
• Easy detection of dust FIR emission through
   inverse K-correction ,
• same flux at 1mm from z 0.5 to 10
• At least ULIRGs 1012 Lo
• ? Numerous 0.1-0.3 per arcmin2
• ? Star Formation Rate SFR gt 100 Mo/yr
• Account for a significant fraction of submm
  background
• Most exceptional HLIRGs 1013 Lo, 1000 Mo/yr
• ? nothing equivalent in the local Universe
  
• Giant starbursts at the peak of star formation,
  z 2-3 ? 1-4, in massive
• proto-elliptical galaxies

4
24µm bright z2 ULIRGs AGN vs
PAHs/starburst Optically faint, 24µm bright
Spitzer sources ? high z sources Large Spitzer
surveys (SWIRE, FLS, etc.) with sensitivity
S24µm250-300µJy (1mJy at z2 ? nLn
1012Lo) Large programs with IRS/Spitzer
spectrometer 15-35µm ? 5-12µm rest Houck et al.
2005, Yan et al. 20052007,etc., Weedman et al.
2006, Farrah et al. 2008, Huang et al. 2008,
etc. - Most of 1mJy sources are AGN dominated
hot dust close to the BH, emitting at 8µm -
Many have strong silicate 9.7µm absorption -Some
have strong PAH features at 6.2, 7.7, 8.6,
11.3µm, or are composite PAH/AGN Such PAH
emission is known to trace strong starbursts (PDR
regions)
5
PAH dominated spectra Yan et al. 2007
Composite AGN-PAH spectra
6
Yan et al. 2007 Starburst vs AGN PAH
spectrum
Starbust
Composite AGN-starburst
7
Stacked spectrum for10 z1.9, Spitzer selected
starburst ULIRGs Huang et al. 2008
8

• Mid-IR properties of SMGs (for comparison)
• Large range of 24µm fluxes
• Average flux rather low
• Pope et al. 2005 (HDFN) ltS24µmgt 240µJy
• Ivison et al. 2007 (SHADES) ltS24µmgt 340µJy
• IRS spectra
• Only for strongest sources
• Pope et al. 2006 ltS24µmgt 380 µJy
• Valiante et al. 2007 ltS24µmgt 500 µJy
• Strong PAH features
• Hints of AGN

9
Selection of z2 starbursts from Spitzer IRAC
24µm photometry IRAC bands allow discrimination
between AGN and starburst (PAH) dominated
sources AGN (and many composites) have more or
less power law IRAC SED Starbursts display a
stellar bump (1.6µm rest) in IRAC
bands Maximum at 5.8µm (4.5µm) band mostly
selects sources at z1.8-2.0 (1.5-1.7) Lonsdale
et al. 2008, Fiolet et al. in prep.
10
Selection of Huang et al. 2008, from 3.6-4.5
vs 3.6-8.0, is practically equivalent to
select z2 starburst
Blue Huang08 Red Fiolet09 Green Lonsdale08
11
How strong in FIR/submm are Spitzer z2
starbursts? Although one expects some
correlation between bright sources in
mid-IR (24µm at z2), FIR and mm/submmfluxes, the
detailed correlation is not obvious Pure AGN
(e.g. Type I QSOs) are known to have flat SEDs
and rather weak mm fluxes This is confirmed
for obscured 24µm-bright Spitzer AGN by
the 1.2mm MAMBO study of Lutz et al. 2005 PAH
dominated (starbursts) (and many composite
sources) have strong/significant mm fluxes, as
confirmed by our MAMBO 1.2mm studies
12

• MAMBO 1.2mm observations of Spitzer z2
  starbursts
• PAH dominated (starbursts) (and many composite
  sources) have
• strong/significant mm fluxes
• Three samples
• Lonsdale, Polletta, Omont et al. 2008 ApJ in
  press
• 61 sources S24µm gt 500 µJy ltS24µmgt820µJy
  5.8µm-peakers
• ? 16 3s detections lt S1.2mm gt 1.49/-0.18
  mJy
• Fiolet, Omont et al. 2009 in prep.
• 32 sources S24µm gt 400 µJy ltS24µmgt540µJy
  5.8µm-peakers
• ? 13 3s detections lt S1.2mm gt 1.54/-0.14
  mJy
• Younger, Omont et al. 2008, submitted to MNRAS
• 12 sources S24µm gt 500 µJy ltS24µmgt800µJy
  3.6-4.5 vs 3.6-8.0
• 9 3s detections lt S1.2mm gt 1.6 /- 0.1
  mJy

13
Large ratio PAH/FIR (1.2M/24µm) emission
Ratio S(1.2mm)/S(24µm) much smaller than most
SMGs whose typical SED is relatively cold,
similar to Arp 220 SED rather similar to M 82
or NGC 6090 But more luminous, without local
equivalent (Desai 07) (or composite AGN IRAS19254)
 Optimized sample 350µm detections
Unbiased sample x Literature sources


14
Far-IR Luminosity and Star Formation
Rate Waiting for Herschel and ALMA, measuring
flux densities between 40µm and 700µm is still
difficult, so that the flux at SED maximum and
LFIR are often uncertain Spitzer has poor
sensitivity at 70 and 160µm ? Exceptionally deep
70-160µm data Huang08, Younger08 Tdust and
LFIR well constrained by data at 70, 160 and 1200
µm Tdust 34 47 K LFIR 2 10 x 1012 Lo ?
Or stack at 70-160µm of gt10 sources
Lonsdale08, Fiolet08 Similar results but more
uncertain ? Or/and measurement at 350µm
(SHARC2/CSO) Kovacs in prep. The few observed
sources yield similar values for Tdust and LFIR
15
Younger et al. 2008
16

• Radio Properties
• Starbursts
• LFIR (and SFR) are known to be strongly
  correlated with radio
• (synchrotron) emission in starbursts
• LFIR/L1.4GHz about constant over several orders
  of magnitude
• from HII regions to ULIRGs
• Spectral index na of starbursts a - 0.7
• AGN
• AGN are known to be even stronger radio emitters
• Radio loud, S1.4GHz gt 300µJy at z2
• Significant radio emission from the AGN even
  below this limit
• Various spectral index and spatial extension

17

• Very deep radio data at 1400 and 610 MHz in a
  0.25-0.5 deg2 field
• Lockman-Owen Field Fiolet et al. in prep.
• MAMBO 1.2mm study of 32 5.8µm-peakers. High
  detection rate
• Combination of radio 1.2mm well discriminate
  AGN and starbursts
• Most of our 24µm sources have
• a radio 30µJy detection
• Ratio 1.2mm/20cm in usual limits
• (e.g. Chapman et al. 2005)
• except 20-35 mm-faint ?5 to 13 AGN?
• A number are radio extended gt 10kpc
• AGN or extended starbursts?
• LFIR inferred from radio/FIR relation

18
CO Study of 24µm bright z2 Spitzer ULIRGs
Reminder Dissecting SMGs through mm CO lines at
IRAM-PdBI
19

• Dissecting SMGs through mm CO lines at
  IRAM-PdBI
• (Very) Large program at the IRAM Plateau de Bure
  millimeter interferometer (PdBI) (Genzel, Ivison,
  Neri, Tacconi, Smail, Chapman, Blain, Cox, Omont,
  Bertoldi, Greve et al.)
• -30 SMGs with z2-3 spectroscopic redshifts from
  radio positions (Chapman, et al.)
• Detection and velocity profiles of CO(3-2) and
  (4-3) lines for 22 SMGs (Neri et al. 2003, Greve
  et al. 2005, Tacconi et al. 2006, Smail et al. in
  prep.).
• Subarcsecond resolution imaging in progress
  (Tacconi et al. 2006, 2008, and in prep.)
• Parallel programs for HST imaging and high
  resolution radio imaging with MERLIN
• Key goals
• - Physical properties and evolution of the SMG
  population
• - How SMGs fit in general picture of galaxy
  evolution and formation

20

• The Plateau de Bure Interferometer

In 2007 PdBI has increased sensitivity by gt2
and baseline by 2 Further gain by 2009 larger
bandwidth 4GHz and more bands (2mm0.85mm).
21
High angular resolution CO mapping at
PdBI Example of mapping CO in an SMG at
PdBI Case of an unresolved 1kpc rotating disk
(2008)
22
Examples of mapping CO in SMGs at PdBI Spatial
and Kinematic Evidence for Mergers
Double or multiple knots, with complex,
disturbed gas motions
Tacconi et al. 2008
23
Current conclusions of PdBI CO survey of SMGs

• High CO detection rate, close to 100 with
  current PdBI sensitivity
• Large fraction are resolved with subarcsecond
  resolution (2/3 are resolved in the radio with
  0.3 MERLIN beam)
• Mm lines of the molecular ISM, are unique to
  trace dynamical masses. (Also large stellar
  masses gt 1011Mo)
• SMGs are short-duration (100 Myr) maximum
  starburst events in the evolution of a major
  gas-rich merger of massive galaxies.
• Different combinations of ordered disk rotation
  and merger driven random motions and inflows
• The high surface densities in SMGs are similar to
  compact quiescent galaxies in the same redshift
  range and much higher than in local spheroids.

24
CO Study of 24µm bright z2 Spitzer ULIRGs
1/2 with upgraded PdBI (in 2007 PdBI has
increased sensitivity by gt2 and baseline by
2) Needs accurate redshift for current PdBI
limited bandwidth 1 GHz for full sensitivity (?
3000 km/s) (? 4 GHz in 2009) PAH determination
of z not currently accurate enough (? 2009)
Optical/NIR redshift measurement is difficult
at z1.8-2.0 especially for starbursts (redshift
desert) Only a dozen of IRS sources have
accurate z (Sajina07, Yan05) We observed 10
with PdBI and detected CO in 9

25

• CO Study of 24µm bright z2 Spitzer ULIRGs 2/2
• We observed 10 sources of Yan07 with PdBI and
  detected CO in 9
• Tacconi et al. in prep., Fiolet et al. in prep.
• PAH dominated sources
• strong CO signals intensity and width comparable
  to SMGs ? MH2
• Most of these sources are 5.8µm-peakers or
  similar sources

26
MIPS16144 Integrated CO 3-2 Emission
strong PAHs strong MAMBO 1.2m flux (2.9mJy) ?
strong CO L. Tacconi in prep.
PAH source, Mambo flux2.93?0.56, z2.13 40 MHz
spectral smoothing, rms0.32 mJy/beam C-configurat
ion
27

• CO Study of 24µm bright z2 Spitzer ULIRGs
  2/2
• We observed 10 sources of Yan07 with PdBI and
  detected CO in 9
• Tacconi et al. in prep., Fiolet et al. in prep.
• PAH dominated sources
• strong CO signals intensity and width
  comparable to SMGs ? MH2
• AGN dominated and composite sources
• weaker (narrower) CO lines
• but most are detected
• CO is detectable with current PdBI sensitivity in
  practically all
• z2 ULIRGs
• High resolution maps in one sources in Winter
  2009?
• Multi-line CO observations in 2mm (1.3mm)
  bands??
• Search in large sample with PAH redshift with 4
  GHz bandwidth??

28
MIPS15949 Integrated CO 3-2 Emission
PdBI Spring08 ROC4
AGN (PAH) source, Mambo flux1.24?0.51,
z2.11 40 MHz spectral smoothing, rms0.28
mJy/beam C-configuration CO43148 CO65222
29
MIPS8327 Integrated CO 3-2 Emission
PdBI Spring08 ROC4
Absorbed source, Mambo flux1.03?0.55, z2.44 40
Mhz spectral smoothing, rms0.24
mJy/beam C-configuration CO54167.52
30
MIPS8342 Integrated CO 2-1 Emission
PdBI Spring08 ROC4
AGN (PAH) source, Mambo flux0.98?0.55,
z1.56 40 MHz spectral smoothing, rms0.19
mJy/beam Note this source was done in
D-configuration CO32
31
Fiolet et al. in prep. Weak 1.2mm MAMBO
Srong 10µm silicate absorption
Broad CO line
Narrow CO line, radio loud
32

• Spatial Extension
• Crucial a significant extension could explain
  stronger PAHs (merger,
• outflow?)
• While most SMGs and AGN are rather compact
  (strong silicate
• absorption)
• But difficult, subarcsec
• Best CO (1.2mm continuum) maps at PdBI of
  strong sources
• ALMA for weak sources
• HST ACS (Huang08), NICMOS (Dasyra,Yan08) many
  sources extended
• but difficult interpretation
• Radio many large extensions (Owen Fiolet in
  prep.)
• but could be AGN jets or merger starbursts

33

• Stellar Masses
• Stellar mass well traced by NIR emission (1.6µm
  bump) of red giants
• IRAC bands at z2
• But risk of contamination by young massive
  TP-AGBs
• Large mass (gt1011Mo) (Berta07? Lonsdale08,
  Fiolet08, Huang-Younger 08)
• Maybe slightly larger than SMGs?

34

• Conclusions. Comparaison with classical SMGs
• PAH luminous z2 sources are strong starbursts
  and belong to the
• SMG family ? special subclass, probably extended
  starburst
• (from late major mergers)
• Compared to the bulk of the SMGs, they have
• Comparable mm/submm fluxes and CO intensities
• Significantly larger 24µm/1.2mm flux ratio
• Probably slightly larger Tdust (mean 40K
  instead 34K)

35
Prospects

• Herschel
• Too small collecting area vs
• ALMA (/500!) for high-z molecules
• But will detect 104s of SMGs in wide
• surveys with full SEDs, LFIR and SFR
• For follow up at PdBI and ALMA
• Will need multi-l data to beat confusion
• SPIRE GTO high-z HERMES 900h mapping most
  Spitzer survey fields
• GO key project ATLAS 500h, 500 deg2 ?
  shallow observations nearby and rare (lensed)
  objects

Herschel bands and SMG SEDs
36
Prospects ALMA
ALMA will provide exquisite CO images, and
multi-line intensities for such galaxies and more
distant ones

• JWST
• MIRI/JWST will have orders of magnitude
  improvements in sensitivity, spatial and/or
  spectral resolution compared with Spitzer
• ? synergy with ALMA

37
(No Transcript)
38
(No Transcript)