MAGMA: The Magellanic Mopra Assessment

1
MAGMA The Magellanic Mopra Assessment

• Tony Wong
• University of Illinois at Urbana-Champaign

Collaborators Annie Hughes (Swinburne/ATNF),
Erik Muller (Nagoya U.), Jorge Pineda (JPL),
Juergen Ott (NRAO), Y. Fukui, A. Kawamura, Y.
Mizuno (Nagoya U.), S. Maddison (Swinburne),
J.-P. Bernard (CESR), Y. Chu, L. Looney (U.
Illinois), C. Henkel (MPIfR), U. Klein (U. Bonn)
2
Motivation

• Magellanic Clouds
• Can obtain both a local (pc-scale) and global
  view of star formation and the ISM
• But, may not be in dynamical equilibrium
• Metallicity lower than Galaxy
• Existing complete ISM surveys
• CO (2.6) NANTEN surveys (Fukui et al. 1999,
  2008)
• HI (1) ATCA Parkes surveys (Stanimirovic et
  al. 1999, Kim et al. 2003)
• H? MCELS (Smith et al.)
• mid-IR S3MC (Bolatto et al.), SAGE (Meixner et
  al.)

3
CO vs. HI in LMC

• Nearest actively star-forming galaxy can easily
  achieve 10 pc resolution with radio telescopes

NANTEN CO
ATCA HI
Fukui et al. 2001, 2008
Kim et al. 2003
4
Why A New Survey?

• 40 pc resolution of NANTEN CO survey leaves many
  molecular clouds unresolved.
• SEST Key Program focused on clouds with bright H?
  emission, and did not employ OTF.

5
MAGMA Components LMC

• Molecular Ridge near 30 Doradus (MX002 2005)
• Properties of GMCs as a function of radiation
  field
• Survey of Clouds in the Inner LMC (M172 2006-7)
• Properties of GMCs across range of environments
• Large scale dynamics and the CO-HI relationship
• Verifying Cloud Masses (M226 2007)
• Compare virial and IR-based methods
• A Complete Flux-Limited Sample (M300 2008-9)
• Properties of the smallest bright CO clouds

6
MAGMA Targets
7
Molecular Ridge

• 120 5 x 5 fields
• 3? sensitivity 200 M? per beam

NANTEN CO
Molecular Ridge
8
CO vs. HI in LMC

• Mopra CO

HI Peak Tb
Eff. lwidth
HI integral
9
CO - HI Offsets

• At 1 resolution, evidence for offsets between CO
  and HI peaks.
• Posibly warm atomic haloes such as seen around
  Galactic GMCs (e.g. Andersson et al. 1991).

10
X Factor

• Assuming clouds are virialized, there is little
  variation in the CO-to-H2 conversion factor as a
  function of ambient FUV field or distance from 30
  Dor.

11
Cloud Mass Spectrum

• Mass spectrum power law index of -1.8 resembles
  that observed in Galactic studies.
• Hint of steeper slope in low-FUV environments.

12
By survey location
By SF activity
Ridge clouds not shown in this panel
LMC log ? (-0.7 0.1) (0.7 0.1) log R
(our data)
MW log ? -0.28 0.55 log R (Solomon ea 1987)
M31 log ? (-0.5 0.3) (0.7 0.2)
(Rosolowsky 2007)
Real variation in amplitude of turbulence between
galaxies?
13
By survey location
By SF activity
Ridge clouds not shown in this panel
Mvir ? Ln
More intracloud medium in larger GMCs, e.g. (e.g.
HI or H2 without associated CO)?
For constant X-factor n 1
MW n 0.8 (Solomon ea 1987)
LMC n 1.2 0.1 (our data)
14
By survey location
By SF activity
Ridge clouds not shown in this panel
As per previous slide, large clouds not as
luminous L ? Rn n lt 2.
For X-factor 3.2 x 1020 cm-2 (K km s-1)-1, ?
170 70 M? pc-2
Similar surface density to MW clouds if X-factor
is (slightly) larger in LMC.
15
Verifying GMC masses in the LMC

• Other methods to measure GMC mass besides virial
  hypothesis
• 1) FIR emission from dust mixed with the
  molecular gas
• 2) Extinction of background stars by molecular
  gas
• 3) mm continuum from cold dust mixed with
  molecular gas
• A three-way comparison between
• XCO mass estimates from MAGMA data
• FIR mass estimate from SAGE 60?m 100?m data
  (method 1)
• Extinction mass estimate from 2MASS 6X Sirius
  data (method 2)
• With Bill Reach (SSC), Jean-Philippe Bernard
  (CESR, Toulouse) Kazuhito Dobashi (Tokyo
  Gakugei)

16
Tracing H2 with FIR Dust Emission

• Method outline
• 70 and 160µm MIPS maps used to estimate dust
  temperature
• Optical depth derived from ratio of modified
  blackbody model at 160µm and actual 160µm map
• Measure dust emissivity per H locally using HI
  map
• Remove optical depth component associated with
  atomic ISM
• Regions of excess optical depth ?
  molecular/ionized gas
• NB Need cold clouds for single dust temperature
  assumption
• (See Reach et al 1994 for full description of
  method)

17
Some example clouds
Contours MAGMA CO data Greyscale IR excess map
18
Tracing H2 with NIR Extinction

• Method outline
• Unlike FIR emission method, atomic and molecular
  ISM should have the same NIR extinction
  properties
• NIR star catalogue (2MASS) used to make map of
  extinction in the LMC (assume Cardelli reddening
  law RV3.1)
• Measure extinction per H locally using HI map
• Remove component associated with atomic ISM
• Regions with excess AV ? molecular/ionized gas
• (See Dobashi et al 2008 for full description of
  method comparison with NANTEN CO data across
  LMC)

19
Example preliminary results
Contours NANTEN CO data Greyscale Excess AV map
From Dobashi et al 2008 NB NANTEN data (not
MAGMA) shown
20
CO vs. HI in LMC
Integrated HI (contour) on Integrated CO
21
CO vs. HI in LMC
Peak HI Temp (contour) on Integrated CO
22
CO vs. HI in LMC
HI vel dispersion (contour) on Integrated CO
23
CO vs. HI in LMC

• Bright CO associated with bright HI, but not vice
  versa.

24
CO vs. Stars in LMC

• CO correlates weakly with both recent and past
  star formation.

25
Molecular to Atomic Gas Ratio
26
CO vs. HI in LMC

• Are molecular clouds formed by colliding of HI
  flows? No correlation of ICO with HI linewidth.

27
CO vs. HI in LMC

• However, CO may be tracing a relatively late
  stage in molecular cloud evolution (Bergin et al.
  2004).

For typical values of v and n, timescale for CO
emission to appear is gt107 yr after shock.
CI/CO ratio may be a much more sensitive probe
of the early postshock gas.
28
Results CO-HI Correlation

• HI necessary but not sufficient for CO detection
• Likelihood of CO detection increases with HI
  integrated intensity and peak brightness.
• Correlation is weak because a lot of bright HI
  emission is not associated with CO emission.
• CO is not associated with enhanced HI linewidth
• No indication of GMC formation from colliding HI
  flows, although subject to interpretation.
• CO/HI ratio not correlated with stellar surface
  density
• Probably limits the role of hydrostatic pressure

29
MAGMA Components SMC
30
MAGMA Components SMC
31
SMC Results

• Size-linewidth relation for northern clouds
  (triangles with errorbars) offset from relation
  for SW clouds (diamonds and solid line).

32
Summary

1. MAGMA will map the brightest CO clouds in the LMC
   and SMC (as detected by NANTEN) at a resolution
   of 45 (11 pc).
2. Maps are revealing molecular cloud properties
   across flux-limited samples in both galaxies.
3. They are also being used to address long-standing
   questions about the ability of CO to trace H2 in
   low-metallicity environments.
4. The relationship between CO and HI, which we are
   investigating globally using the NANTEN data, can
   be studied on the scales of individual clouds.