Molecular Gas and Star Formation in Dwarf Galaxies

1
Molecular Gas and Star Formation in Dwarf Galaxies

• Alberto Bolatto
• Research Astronomer
• UC Berkeley
• Adam Leroy
• Josh Simon
• Leo Blitz

Hard working grad students
2
Why should you care?

• Extreme properties are often sought for in
  Astronomy as one way to sharpen our understanding
  of fundamental concepts
• Dwarf galaxies
• are the first structures to form in bottom-up
  ?CDM cosmologies
• have low heavy element abundances, just like
  primordial systems
• are the simplest systems
• Local dwarfs are windows onto the high-z Universe

3
A single-dish/interferometric survey

• MIDGet
• A CO survey of IRAS-detected, compact, nearby,
  northern dwarf galaxies out to VLSR1000 km s-1,
  with rotational velocities under 100 km s-1
• Observed 121 central pointings with the Kitt Peak
  12m
• Follow up of 30 galaxies mapped using BIMA
• Fabian Walters OVRO sample

UASO 12m
BIMA
4
Two questions

• What global properties distinguish galaxies with
  and without CO?
• Some of the best molecular gas predictors are
  surprising LK, Mdyn/LK, B-K (B-V)
• Are there any differences between large and dwarf
  galaxies in their molecular gas/star formation
  properties?
• Remarkably very few, even where some where
  expected

5
Distributions of detections/nondetections

• Best predictors of CO LK, LB, Hubble Type,

1/5 Z?
6
Distributions of detections/nondetections

• Best predictors of CO LK, LB, Hubble Type, FIR
  luminosity, B-K color, K-band mass to light ratio

7
One of the best predictors of CO in the survey

• M/L 3 (B-band) and 2 (K-band)
• But the correlation is much tighter at the low
  end in B light CO nondetections are
  systematically fainter in K-band!

8
What is the driving relationship?

• LFIR, LK, LB, B-K, Hubble Type, Z, are all
  correlated
• Can we identify a driving parameter?
• Normalizing by LK removes trends and minimizes
  dispersion

M
9
What is the driving relationship?

• LFIR, LK, LB, B-K, Hubble Type, Z, are all
  correlated
• Can we identify a driving parameter?
• Normalizing by LK removes trends and minimizes
  dispersion
• Mmol/LK is the tightest correlation. Across all
  galaxy sizes Mmol/LK0.075

10
What does it mean?

• Facts
• Tightest Mmol correlation is with LK, a proxy for
  M and S
• Correlations with Mgas (HI) or Mdyn are
  considerably weaker
• Taken together, suggest that what matters in the
  HI?H2 conversion is the amount of matter in the
  disk (S), not just the amount of stuff
• Correlations with B-K could arise from enhanced
  photodissociation/less dust in bluer systems
• but systems with no CO tend to be underluminous
  (for their mass) in K-band, not overluminous in
  B-band
• Suggests that photodissociation plays only a
  secondary role in setting the global amount of H2
  
• This is indirect evidence in support of the local
  density (pressure) controlling HI?H2

11
Are large and dwarf galaxies different in their
molecular gas/star formation properties?
12
The SFR vs. H2 relationship

• 1.4 GHz flux traces star formation (e.g., Condon
  et al. 2002, Murgia et al. 2002
  SF?SN?CR?synchrotron?)

13
The SFR vs. H2 relationship

• 1.4 GHz flux traces star formation (e.g., Condon
  et al. 2002, Murgia et al. 2002
  SF?SN?CR?synchrotron?)
• MIDGet and large galaxies fall on the same
  ?SFR-?H2 correlation

14
The SFR vs. H2 relationship is independent of Z!

• 1.4 GHz flux traces star formation (e.g., Condon
  et al. 2002, Murgia et al. 2002
  SF?SN?CR?synchrotron?)
• MIDGet and large galaxies fall on the same
  ?SFR-?H2 correlation using the Galactic Xco!

15
Attempts to correct CO-H2 for metallicity fail

• There is no segregation by inferred metallicity
  (using Richer McCall 1995)

16
Attempts to correct CO-H2 for metallicity fail

• There is no segregation by inferred metallicity
  (using Richer McCall 1995)
• Corrections destroy the agreement!

17
Ways out of a constant Xco

• Size-dependent corrections to RC-SFR (e.g. Bell
  2003)?
• Even then large changes in Xco are out of the
  question
• A different SFR-H2 regime for dwarf galaxies?

18
The sweet spot for star formation efficiency

• A maximum star formation efficiency at 1010 M??
• To a first approximation galaxy-size /
  metallicity corrections to LFIR and Xco cancel
• A large Xco(Z) makes the maximum more pronounced

19
Summary

• Mmol correlates very well with LK, not with MHI
  or Mdyn
• Indirect support for a local density/pressure
  controlled HI?H2 transition
• Same SFR-H2 relationship for dwarfs and large
  galaxies, suggesting constant CO-H2 for star
  forming gas despite changing metallicity
• A minimum H2 depletion time / maximum SF
  efficiency at 1010 M??

20
CARMA is moving forward