Properties of Dusty, Highly-Luminous Starbursts Back to the First - - PowerPoint PPT Presentation

SMG20 – Twenty Years of Submillimetre Galaxies August 02, 2017

Dominik A. Riechers

Cornell University

Properties of Dusty, Highly-Luminous Starbursts Back to the First Billion Years of Cosmic Time

§ AzTEC-3: Most Distant Massive Starburst Galaxy 2010-13 § Mgas = 5.3 x 1010(aco/0.8) Msun SFR: ~1100 Msun/yr …or >3 Msun/day § Early Galaxy Proto-Cluster: 11 star-forming galaxy companions within r~2 Mpc

2x2 arcmin2

Molecular Gas

COSMOS/AzTEC-3 (z=5.3)

§ AzTEC-3: Most Distant Massive Starburst Galaxy 2010-13 § Mgas = 5.3 x 1010(aco/0.8) Msun SFR: ~1100 Msun/yr …or >3 Msun/day § Early Galaxy Proto-Cluster: 11 star-forming galaxy companions within r~2 Mpc

2x2 arcmin2

Molecular Gas

COSMOS/AzTEC-3 (z=5.3)

§ 2017:

• 17 z>5 DSFGs (incl. 3 at z>6)
• 1 ALMA (serendip.; 1 mm)
• 1 AzTEC 1.1 mm
• 2 SCUBA 850 µm
• 6 SPT 1.4+2.0 mm
• 7 Herschel/SPIRE “Red” 250-500 µm
• 11 strongly lensed
• 3-4 weakly magnified
• 2-3 unlensed
• 15/17 from “blind” CO redshifts

Riechers et al. 2010, 2013, 2014, 2017abc*; Capak et al. 2011; Walter et al. 2012; Combes et al. 2012; Vieira/Weiss et al. 2013; Strandet et al. 2016, 2017; Pavesi et al. 2017*; Fudamoto et al. 2017; Zavala et al. 2017

§ Most common UV/optical lines typically dust-obscured in DSFGs § DSFGs have rich (sub)millimeter/far-infrared spectra, even at z>6 Þ Main diagnostic tools to understand physical properties of intense star formation

Riechers et al. 2013

§ Probing deeper/broader by utilizing stacked spectra of lensing-magnified sources Þ powerful tool to access fainter diagnostic lines … but what can we learn from these lines?

Spilker et al. 2014; Zhang et al. 2017 250 - 7000 GHz

cm/mm: rich in line + continuum diagnostics

CO ‘ladder’

total gas masses excitation, dynamics

• phys. conditions

CNO fine structure lines

ISM gas coolant

Synchrotron + Free-Free

(AGN+SNR) star formation

Thermal dust

(young stars) star formation

JVLA ALMA+CCAT

PAHs + SiL

SPICA JWST

Smail et al. 2011 Swinbank et al. 2011

Cosmic Eyelash model SED

*model*

cm/mm: rich in line + continuum diagnostics

CO ‘ladder’

total gas masses excitation, dynamics

• phys. conditions

CNO fine structure lines

ISM gas coolant

Synchrotron + Free-Free

(AGN+SNR) star formation

Thermal dust(young stars)

star formation

JVLA ALMA+CCAT SPICA JWST

Smail et al. 2011 Swinbank et al. 2011

Cosmic Eyelash model SED

*model*

PAHs + SiL

Riechers et al. 2014a

6.2 µm PAH

§ >200 CO detections in DSFGs to date:

• ~50 SCUBA + AzTEC + LABOCA (Bothwell+; Riechers+; Coppin+; Yun+; Huynh+, …)
• ~100 Herschel (Harris+; Lupu+; Riechers+; Zavala+; Fudamoto+, …)
• ~40 SPT (Weiss/Vieira+; Strandet+; Aravena+)
• ~5 ACT (Su+; Roberts-Borsani+; …)
• ~12 Planck (Canameras+; Harrington+)
• several “misc” (Swinbank+; Lestrade+; Leung+; Tamura+; Karim+, …)

Carilli & Walter 2013

starbursts starbursts

log LFIR ~ proxy for SFR log L’CO ~ proxy for M(H2)

m a i n s e q u e n c e

MS galaxies

Carilli & Walter 2013 ARAA; after Daddi et al. 2010, Genzel et al. 2010

Simplest Version of “Star Formation Law”: Spatially Integrated Observables

L’CO vs. LFIR

as a surrogate for

Mgas vs. SFR

One super-linear relation or Two sequences (quiescent/starburst)

Bimodal or running conversion factor

…many subtleties, but: High-z galaxies higher on both axes Quiescent and Starburst Galaxies

Aravena et al. 2016 (see also Bothwell et al. 2013; Magdis et al. 2012)

• Star-forming galaxies show 10-30x higher gas fractions at z=1-3 compared to present day
• Increase in gas fraction likely dominantly responsible for increase in star formation history
• Little evidence for significant redshift evolution in gas depletion timescales in starbursts

Þ Star formation activity is elevated, but underlying physics are similar const.?

Riechers et al. (2011d, 2011f); see also: Ivison et al. (2011)

Submillimeter Galaxies: Gas-Rich Starbursts along the “Merger Sequence” at z>2? Ø Nearby major mergers show increased SF efficiency relative to disks Ø SMGs could be “scaled up” versions of nearby IR-luminous galaxies/mergers Line profile and gas dynamics sometimes show evidence for major mergers

Early stage

~30kpc & 750km/s separation

Intermediate stage

~20kpc & <100km/s separation

Late stage

7-15kpc nucleus & tidal structure single broad, multi-peaked line abundant low-excitation gas

SDP-81 (z=3.0)

Hodge et al. 2012 ALMA collab. et al. 2015; Dye et al. 2015; Swinbank et al. 2015; see also Rybak et al. 2015; Hatsukade et al. 2015

High-Resolution Molecular Line Spectroscopy w/ interferometers yields velocity fields Gas Kinematics: Evidence for 2-15 kpc, rotating gas disks in DSFGs

Riechers et al. 2013b

HFLS3 (z=6.34):

• smooth velocity profile at 0.2” res.
• high velocity dispersion across 3.5 kpc
• superposition of two components

with very different line widths Þ likely disturbed/interacting system

Riechers et al. 2014b

AzTEC-3 (z=5.30):

• Little evidence for velocity gradient at 0.6” res.
• high velocity dispersion across 4 kpc
• Simple, broad symmetric line profile
• nearby UV-bright companions

Þ likely disturbed/interacting system

[CII]

1000 km/s

§ 4 HCN & 4 HCO+ & 2 HNC detections in DSFGs to date:

• SMM J2135-0102 (z=2.3; HCN, tentative; Danielson et al. 2011)
• SDP 9 & 11 (z~1.7; 2x HCN, 2x HCO+, 1x HNC; Oteo et al. 2017)
• ACT J0210+0015 (z~2.6; HCN, HCO+, HNC; Riechers+)
• Planck G092.5+42.9 (z~3.3; HCO+; Riechers+)

also: several S/N~3 features in SPT stack (Spilker et al. 2014)

Carilli & Walter 2013

Example: HFLS3 (z=6.34): 7 H2O lines detected up to Eupper ~ 450 K, (shallow) coverage of 22 lines up to Eupper > 1200 K Lines seen in emission, approaching strengths comparable to the brightest CO lines Þ Given the large difference in ncrit to CO, significant collisional excitation is unlikely Þ H2O excitation is indistinguishable from Arp 220 Þ Consistent with radiative excitation in warm, obscured starburst Riechers et al. 2013b

Linear relations:

• H2O lines dominated by FIR pumping
• H2O traces dense, warm gas at high tdust

Yang et al. 2016

§ >60 [CII] detections in DSFGs to date § (much) smaller samples for [OI]63, [OIII]88, [NII]122 & 205 [CII] associated with multiple ISM phases (neutral/ionized gas, star forming regions…) à need multiple lines to study ISM actual properties

Carilli & Walter 2013

CI à CII @11.26 eV à [CII] from PDR+HII HI à HII @ 13.6 eV NI à NII @14.53 eV à [NII] from HII only [CII] 158µm & [NII] 205 µm have similar critical density in HII regions à [CII]/[NII] ratio gives fraction of [CII] from neutral vs. ionized ISM à [NII]122/[NII]205 gives HII region density

Observed: [CII], [NII]122, [NII]205 Have:

• [CII] fraction from PDRs
• HII region density

[OI] from PDRs à [CII]/[OI] sensitive to PDR density

• CLOUDY models:

High density and radiation intensity à high [CII]/[NII]

• Similar ratio in nearby ULIRGs,

dusty LBGs, DSFGs, QSOs à [NII] coming from diffuse (non-starbursting) gas

• high [CII]/[NII] in dust-poor

LBGs à similar to nearby low-metallicity dwarfs

Updated from Pavesi, DR et al. 2016

HIGH [CII] from ionized gas LOW [CII] from ionized gas

metallicity metallicity ~Tdust

Cormier et al. 2015

[CII]/[NII] not very sensitive to Tdust & lines have similar ncrit in HII regions [CII]/[NII] appears to rise steeply with metallicity à Fraction of [CII] from HII regions seems to decrease à Low [NII] thus may indicate that most N (and C, O) in higher ionization state à Consistent with rising [OIII]/[NII] & [OIII]/[CII] with metallicity

Assuming [CI] is optically-thin and has a standard abundance: ~2-3x higher inferred M(H2) than with standard aCO à either CO underpredicts gas mass, or C has higher than nominal abundance

Bothwell et al. 2017 (see also Weiss et al. 2005; Riechers et al. 2009)

à M. Bothwell’s talk

ALMA Herschel/SPIRE

• idea: z>4 galaxy dust SEDs preferentially peak beyond 500 µm

Þ can use (sub)mm colors to determine reasonable photometric redshifts Þ“red” sources are strong candidates for starbursts at the earliest epochs

wavelength 250 350 500 µm Flux density

Spectral Energy Distribution M100

Credits: X-ray: NASA/CXC/SAO/D.Patnaude et al, Optical: ESO/VLT, Infrared: NASA/JPL/Caltech

Riechers et al. 2013b

SPT vs. Herschel-Red selection (CO spec-z only):

• strongly overlapping, but different redshift distributions
• indistinguishable when only taking “red” SPT sources

Þ perhaps not surprising: preferentially followed up 870µm/1.3mm-bright Herschel-Red sources Þ Combine samples to obtain median redshift: zmed=4.42

2 4 6 8 10 2 3 4 5 6 7 N redshift SPIRE-Red SPT 2 4 6 8 10 2 3 4 5 6 7 N redshift SPIRE-Red SPT-Red 2 4 6 8 10 12 14 2 3 4 5 6 7 N redshift SPIRE+SPT-Red

Riechers et al. 2017a Beyond Herschel-Red Selection: Red source(s) w/ SEDs rising to 870 µm Only 1/300 red sources followed up over ~1000 deg2 w/ SCUBA-2/LABOCA: Unlensed Hyper-LIRG: 25 mJy @870 µm

10-1 100 101 102 103 104

λobs [µP]

10-3 10-2 10-1 100 101 102

)OX[ DHnsLty [P-y] JHnHrDO RStLFDOOy thLn JHnHrDO (12µP-3PP RnOy) Herschel/S3I5( A/0A A3(X//AB2CA

10-2 10-1 100 101 102 103

λrest [µP]

10-1 100 101 102 103 104

λrest [µP]

10-4 10-3 10-2 10-1 100

)OX[ [nRrDP]OLHd Dt rHst-IrDPH 75 µP]

AD)S-27 JHnHrDO H)/S3 (yHODsh ArS220 A/(SS AD)S-27

Riechers et al. 2017a ALMA: Binary HyLIRG (separation: 9 kpc; SFR: ~2500 Msun/yr) at z~5.7 2 kpc diameter “maximum starbursts”: SSFR ~ 750 Msun/yr/kpc2

n Rich ISM diagnostics reveal the physical properties and chemical

composition of the gas

n Gas dynamics reveal a mix of rotating disks and dispersion-dominated,

likely interacting systems (which commonly are “maximum starbursts”) Q: How many how bright DSFGs at what redshift are required to really put current models of their formation to rest?

n Significant progress made on uncovering the high-z tail of DSFGs

extending into the first billion years after the Big Bang

n The high star formation rates in DSFGs compared to nearby ULIRGs

are dominantly driven by an increased ISM content/high gas fractions

http://www.astro.caltech.edu/alma2017/

• Serena Viti
• Adam Leroy