Dust/Gas Correlation in the Large Magellanic Cloud: New Insights from the HERITAGE and MAGMA surveys Julia Roman-Duval July 14, 2010 HotScI.

Dust/Gas Correlation in the Large Dust/Gas Correlation in the Large Magellanic Cloud:Magellanic Cloud:

New Insights from the HERITAGE New Insights from the HERITAGE and MAGMA surveysand MAGMA surveys

Julia Roman-DuvalJulia Roman-DuvalJuly 14, 2010July 14, 2010

HotScIHotScI

Why care about dust and gas in the ISM ?Why care about dust and gas in the ISM ?

• Constrain galactic evolution models

SFR g1.4-1.5 (Schmidt 1959, Kennicutt 1998), SFR H2 (Bigiel et al.

2008, Leroy et al. 2008)

• Dust shields molecular clouds from interstellar radiation field, allows molecules to form and gas to cool, and SF to proceed

• We can observe dust easily (emission of absorbed stellar light in FIR)

• H2 does not have a dipole moment = no emission at gas temperature

• Use CO as a tracer of H2, problem in low metallicity galaxies

OBJECTIVEOBJECTIVE1. Provide a brief summary of the effects of

metallicity on dust/gas correlation and the structure of MCs

2. CO-dark molecular gas problem

3. Present current and new data sets that will shed more light on the issue

4. Present preliminary results for the LMC derived from Herschel data from the Science Demonstration Phase (SDP)

Dust, CO, H2

Dust, H2, C, C+

Dust, H, C+

Wolfire et al. (2010)

Unlike CO, HUnlike CO, H22 self-shielded due to self-shielded due to

absorption lines in the UVabsorption lines in the UV

Interstellar Radiation Field (ISRF) attenuated by DUST

dust = gas/GDR

Metallicity EffectsMetallicity Effects

CO fraction determined by AV (photo-dissociation)

Glover et al. (2010)

H2 fraction determined by nZ (formation timescale)

Metallicity EffectsMetallicity Effects

Bolatto et al. (1999)

Expect a significant fraction of molecular gas to be invisible to CO observations in low metallicity galaxies

H0/Z = thickness of the C0 regionH0/Z = thickness of the C0 region

AV Z for a given gas mass

FIR Excess from SPITZER FIR Excess from SPITZER Observations of the LMCObservations of the LMC

Bernard et al. (2008)

Correlation between FIR emission from dust and NH expected from a constant GDR measured in diffuse regions where no molecular gas is expected to exist

Excess of FIR emission compared to the measured gas column (from HI 21 cm and CO emission) at high columns

=> CO dark molecular gas ?

Nx = GDR Ndust - N(HI) - N(H2

CO)

FIR excess in the LMCFIR excess in the LMC

FIR excess correlated with high density regions

FIR excess and CO-dark FIR excess and CO-dark molecular gasmolecular gas

1. Is the FIR excess due to CO dark molecular gas ?

2. Can dust/FIR measurements trace this CO-dark molecular gas ?

DATA setsDATA sets

• LMC (Z = 0.5 Zo), SMC (Z = 0.2 Zo) closest low-Z galaxies

• Pre-Herschel data sets from the Surveying the Agents of Galactic Evolution (SAGE) project (Meixner et al. 2006):– SPITZER/MIPS (24-160 m, 40” resolution)– SPITZER/IRAC (3-8 m)– IRAS (12-100 m, 4.3’ resolution)– NANTEN 12CO (J = 1-0) (2.6’ resolution)– ATCA+Parkes HI 21 cm (1’ resolution)

• Limited by IRAS resolution (63 pc at the distance of the LMC) to estimate dust SED and dust/gas surface density

• Width of the CO dark molecular gas region (AV = 0.7)– L = AVN0/(<n>Z) = 0.8 pc @ n= 1000 cm-3

– L = 8pc @ n = 100 cm-3

• Need higher resolution to better resolve the structure of molecular clouds envelopes

New data setsNew data sets FIR

• Herschel HERITAGE (HERschel Inventory of The Agents of Galactic Evolution)

• Resolution limit: SPIRE 500 (40”)

• Images of the entire LMC and SMC at 100, 160 (PACS), 250, 350, 500 (SPIRE) m.

12CO

• MAGMA (MAGellanic MOPRA Assessment (12CO follow up on molecular clouds detected by NANTEN)

• 40” resolution, 0.5 K km/s sensitivity

HerschelHerschel SDP: Dust surface density SDP: Dust surface density

QuickTime™ and a decompressor

are needed to see this picture.QuickTime™ and a decompressor

are needed to see this picture.

Gordon et al. (2010)

Meixner et al. (2010)

IRAS 100 m resolution (4.3’)dust (IRAS+MIPS) (HI) (ATCA+Parkes HI 21 cm)(H2

CO) NANTEN 12CO

1’ resolution dust (MIPS + HERSCHEL) (HI) (ATCA+Parkes HI 21 cm)(H2

CO) NANTEN 12CO (2.3’ resolution)

NT80 at 1’ resolution dust (MIPS + HERSCHEL) (HI) (ATCA+Parkes HI 21 cm)(H2

CO) MAGMA 12CO

NT80 at 4.3’ resolutiondust (IRAS+MIPS) (HI) (ATCA+Parkes HI 21 cm)(H2

CO) NANTEN 12CO

NT80

IRAS 100 m resolution (4.3’)dust (IRAS+MIPS) (HI) (ATCA+Parkes HI 21 cm)(H2

CO) NANTEN 12CO

1’ resolution dust (MIPS + HERSCHEL) (HI) (ATCA+Parkes HI 21 cm)(H2

CO) NANTEN 12CO (2.3’ resolution)

NT80 at 4.3’ resolutiondust (IRAS+MIPS) (HI) (ATCA+Parkes HI 21 cm)(H2

CO) NANTEN 12CO

NT80 at 1’ resolution dust (MIPS + HERSCHEL) (HI) (ATCA+Parkes HI 21 cm)(H2

CO) MAGMA 12CO

NT71

Dust/Gas correlationDust/Gas correlation

Roman-Duval et al. (2010)

Dust/gas spatial correlationDust/gas spatial correlation

Roman-Duval et al. (2010)

NT80dust (1st panel)GDRdust/((HI) + (H2

CO))--- (HI)--- (H2

CO) Regions of FIR excess

Dust/gas spatial correlationDust/gas spatial correlation

Roman-Duval et al. (2010)

NT71dust (1st panel)GDRdust/((HI) + (H2

CO))--- (HI)--- (H2

CO) Regions of FIR excess

Excess of FIR emission compared to the observed gas surface density near the envelopes of MCs consistent with CO dark molecular gas

Variations of XVariations of XCO CO with Awith Avv

Glover et al. (2010)

€

XCO =N(H2)

WCO

Transition between CO core and CO-free H2 envelope

Variations of XVariations of XCO CO with Awith Avv

Roman-Duval et al. (2010)

Preliminary results

€

XCO =GDR Σdust − Σ(HI)

WCO

• Emissivity variations– Grain coagulation in dense, molecular regions

(Paradis et al. 2009)

• Gas-to-dust ratio variations– Dust destruction in the diffuse ISM by shocks

(Jones et al. 1996)– Grain growth in the dense molecular phase

Other possible causes for observed Other possible causes for observed deviations in the dust/gas correlationdeviations in the dust/gas correlation

Emissivity

GDR

NT80 NT71

Assumptions

Dust temperature Dust temperature in ISM phasesin ISM phases

NT80SFR = 0.018 Mo/kpc2/yr

NT71SFR = 0.042 Mo/kpc2/yr

Ongoing/future workOngoing/future work

• Coming soon: full HERITAGE Mosaics of the LMC and SMC in PACS100, PACS160, SPIRE250, SPIRE350, SPIRE500 m

• Epoch 1 has been reduced, scientific analysis under way

• Include H2 radiative transfer, cooling and heating, and basic H2 and CO chemistry in DIRTY radiative transfer code

• Extend the SDP analysis to larger sample of MCs in the SMC, where metallicity effects are more important

Conclusion• Molecular cloud envelopes in low metallicity galaxies (e.g., LMC, SMC)

galaxies probably hide large amounts of molecular gas not traced by CO

• Deviations in the dust/gas surface density correlation (FIR excess) are not likely to be caused by gas-to-dust ratio or emissivity variations between the diffuse and dense phases

• Envelopes of H2 not traced by CO are more likely

• Dust emission in the FIR is potentially a good tracer of this CO-dark molecular gas (see also, Isarel et al. 1997, Leroy et al. 2007, 2009)

• The dust temperature is lower by a few degrees In the molecular phase compared to the diffuse phase

Happy Bastille Day !