Dust-to-Gas Ratios and XCO in Nearby Galaxies with HERACLES

Transcription

1 Dust-to-Gas Ratios and XCO in Nearby Galaxies with HERACLES Karin Sandstrom (MPIA) Adam Leroy (NRAO), Fabian Walter (MPIA), Andreas Schruba (MPIA) and the HERACLES team April 10, 2010 Stars to Galaxies, Gainesville FL

2 igure 8. Estimate of the CO-to-H 2 conversion factor X CO,est, Figure 9. Estimate of the CO-to-H 2 conversion factor X C Karin Sandstrom - DGR & XCO from HERACLES Motivation H2 is difficult to directly observe under GMC conditions. We rely on tracers, and CO is often the most convenient. The CO-to-H2 conversion factor XCO is expected to be a function of environment. No. 2, 2007 Our knowledge of DGR is closely tied to XCO. DUST MASSES IN SINGS GALAXY SAMPLE 885 Glover & Mac Low 2010 Draine et al. 2007

3 Definitions CO-to-H2 conversion factor XCO (cm -2 (K km s -1 ) -1 ) XCO = N(H2)/ ICO H2 column density (cm -2 ) CO J=(1-0) integrated intensity (K km s -1 ) Dust-to-Gas Ratio DGR = ΣD/(ΣHI + ΣH2) Dust Mass Surface Density (M /pc 2 ) } Gas Mass Surface Density (M /pc 2 ) ΣH2 = αxco ICO α converts N(H2) to mass units XCO,MW ~ cm -2 (K km s -1 ) -1 DGRMW ~ 0.01

4 Observational Constraints on XCO Young & Scoville (1991) Virial mass of MC from CO line widths comp. with H2 masses. XCO,MW ~ (Young & Scoville 1991) Issues with angular resolution - need to measure cloud size.

5 Observational Constraints on XCO Strong & Mattox (1996) Virial mass of MC from CO line widths comp. with H2 masses. XCO,MW ~ (Young & Scoville 1991) γ-ray emission trace gas mass (interactions of CR with gas). XCO,MW = 1.9± (Strong & Maddox 1996) γ-ray XCO,MW = (Abdo et al. 2010) Only possible in MW and very nearby galaxies. Requires a good model for the CR distribution.

6 Observational Constraints on XCO Fix DGR = ΣD/(ΣHI +αxco ICO) Solve for observed ICO DGR predicted ICO Virial mass of MC from CO line widths comp. with H2 masses. XCO,MW ~ (Young & Scoville 1991) γ-ray emission trace gas mass (interactions of CR with gas). XCO,MW = 1.9± (Strong & Maddox 1996) XCO,MW = (Abdo et al. 2010) Dust as a tracer of total gas column. Dame et al. (2001) XCO,MW = 1.8± (Dame et al. 2001)

7 Using Dust to Trace Total Gas Column Leroy et al. (2009) Fix DGR = ΣD/(ΣHI +αxco ICO) Solve for ΣH2 from far-ir Fix DGR based on what is observed nearby a cloud, measure dust, H I and CO and derive XCO. Currently possible in MW, LMC and SMC due to angular resolution. ICO (K km s -1 ) Figure 7. FIR Σ FIR see also: Israel (1997), Leroy et al. (2007)

8 Using Dust to Trace Total Gas Column Boselli et al Fix DGR = ΣD/(ΣHI +αxco ICO) Solve for Fix DGR based on an assumption of metallicity dependence, measure dust, H I and CO and derive XCO. also: Regan (2000)

9 Using Dust to Trace Total Gas Column In the ideal case we would not have to make any assumption about the DGR - instead constrain both independently. Solve for DGR = ΣD/(ΣHI +αxco ICO) Solve for This is what we ll attempt to do in the following.

10 The Sample HERACLES: Survey of CO J=(2-1) in nearby galaxies using HERA on the IRAM 30m (previous talks by Leroy & Bigiel). Leroy et al. (2009) NGC 0628 Red=160 μm, Green = CO, Blue = H I NGC galaxies in the HERACLES sample overlaps THINGS and SINGS/LVL. THINGS - Walter et al. (2008) SINGS - Kennicutt et al. (2003) LVL - Dale et al. (2009) Throughout we assume (2-1)/(1-0) ratio of 0.8. (Leroy et al. 2008) All maps convolved to 40 MIPS 160 resolution.

11 Technique We d like to constrain both DGR and XCO with the fewest assumptions possible. DGR = ΣD/(ΣHI +αxco ICO)

12 Technique We d like to constrain both DGR and XCO with the fewest assumptions possible. DGR = ΣD/(ΣHI +αxco ICO) MIPS 160 μm NGC 0628 ΣD from Draine et al. (2007) model fits to MIPS 24, 70 & 160 μm bands (need 70/160 for TD)

13 Technique We d like to constrain both DGR and XCO with the fewest assumptions possible. DGR = ΣD/(ΣHI +αxco ICO) MIPS 160 μm HERACLES CO NGC 0628 ΣD from Draine et al. (2007) model fits to MIPS 24, 70 & 160 μm bands (need 70/160 for TD)

14 Technique We d like to constrain both DGR and XCO with the fewest assumptions possible. DGR = ΣD/(ΣHI +αxco ICO) MIPS 160 μm HERACLES CO THINGS H I NGC 0628 ΣD from Draine et al. (2007) model fits to MIPS 24, 70 & 160 μm bands (need 70/160 for TD)

15 Technique DGR = ΣD/(ΣHI +αxco ICO) 1) Pick out regions where we are confident H I and CO contribute to Σgas. 2) Assume these regions can be described by a single value of DGR and XCO. 3) Find the XCO that minimizes the fractional scatter in DGR. } } } Future work using CO upper limits may expand this, but for now stick to CO detections. Probably only valid when there is not a large gradient in physical conditions (metallicity, etc) Similar to maximizing correlation between dust and gas surface densities, finding best fit line, etc.

16 NGC 5055 Example Error bars from bootstrapping. Fractional scatter = ΔDGR/DGR Red line represents 1-σ contraint on XCO.

17 NGC 5055 Example Error bars from bootstrapping. Fractional scatter = ΔDGR/DGR Red line represents 1-σ contraint on XCO.

18 NGC 5055 Example Error bars from bootstrapping. Fractional scatter = ΔDGR/DGR Red line represents 1-σ contraint on XCO.

19 NGC 5055 Example Error bars from bootstrapping. Fractional scatter = ΔDGR/DGR Red line represents 1-σ contraint on XCO.

20 NGC 5055 Example Error bars from bootstrapping. Fractional scatter = ΔDGR/DGR Red line represents 1-σ contraint on XCO.

21 NGC 5055 Example Error bars from bootstrapping. Fractional scatter = ΔDGR/DGR Red line represents 1-σ contraint on XCO.

22 NGC 5055 Example Error bars from bootstrapping. Fractional scatter = ΔDGR/DGR Red line represents 1-σ contraint on XCO.

23 NGC 5055 Example Error bars from bootstrapping. Fractional scatter = ΔDGR/DGR Red line represents 1-σ contraint on XCO.

24 Results NGC 0628

25 Results NGC 5055

26 Results NGC 3184

27 Results NGC 3198

28 Results Mean XCO = (c.f. Draine et al. 2007) NGC 4214 NGC 3198 Variations of an order of magnitude in XCO and DGR across the sample. NGC 0628 MW Mean = 3.7 XCO and DGR inversely correlated. Limits only: NGC 2976, NGC 0925, NGC 3077 NGC 4736 Lowest DGR/highest XCO in NGC irregular galaxy Highest DGR/lowest XCO in NGC nuclear starburst ring dominated?

29 Results NGC 4736 DGR varies with metallicity. NGC 0628 NGC DGR = DGRMW (O/H)/(O/H)MW 2 Metallicity trend is clear even over limited range of (O/H). Generally consistent (within factor of ~2) of directly scaling with (O/H). Metallicities from Pilyugin et al. (2004) (mean over region of galaxy used in the analysis)

30 Results NGC 3198 No clear evidence for metallicity dependence over this range of (O/H). XCO = NGC 0628 XCO = XCO,MW XCO = XCO,MW (O/H)MW/(O/H) NGC 4736 Selecting regions where we see CO probably yields a biased picture of the full range of XCO variation.

31 Summary For the 18 galaxies with CO from HERACLES, H I from THINGS and far-ir observations of SINGS/LVL we constrain DGR & XCO without prior assumptions about their environmental dependence. Mean XCO for the sample is ~ Highest XCO is observed in Irr galaxy NGC 4214 and lowest XCO in NGC 4736 which has a bright circumnuclear starburst ring. DGR consistent with being proportional to (O/H). DGR and XCO have an inverse correlation with large scatter. Future work using HERACLES observations + KINGFISH will allow us to expand parameter space of metallicity & DGR where XCO can be constrained.