Dust distribution and properties in dwarf galaxies

1
Dust distribution and properties in dwarf galaxies

• Ute Lisenfeld
• Universidad de Granada
• Collaborators
• Monica Relaño (Institute of Astronomy, Cambridge)
• José Vílchez (Instituto de Astrofísica de
  Andalucía, Granada)
• Israel Hermelo (Universidad de Granada)
• Eduardo Battaner (Universidad de Granada)

2

Goals

• Study dust via extinction and emission in HII
  regions of starbursting dwarf galaxies Dust/ISM
  is expected to be affected by strong radiation
  field and energy input
• Dust extinction from Balmer decrement
• HST images for H? and H? ? High resolution
  allowing to study the small scale variation of
  the dust
• Large scale extinction structure
• Dust emission from Spitzer and SCUBA
• Information on warm and cold dust
• Compare emission and extinction ? information
  about the dust distribution
• Study use of dust emission at 24?m as a SF tracer
  as a function of metallicity.

3
Distribution of the Extinction and Star Formation
in NGC 1569

• Nearby dwarf galaxy (D 2.2 Mpc) hosting two
  SuperStarClusters (SSC A and B) near the center
  of the galaxy
• Good object to study the distribution of the
  dust The violent star formation has strongly
  influenced the distribution of the gas in the
  galaxy.

• Global burst of gt 0.1Gyr ended 5-10Myr ago, in
  which the SSCs A and B are formed (Greggio et al.
  1998)
• The strong stellar winds and SN explosions from
  SSCs A and B have swept up the interstellar
  ionized gas surrounding the star clusters
  (González-Delgado et al. 1997).
• Martin (1998) has identified spectroscopically an
  expanding shell centered at the location of SSCs.
  

A
B
Continuum-subtracted Ha image of NGC 1569
(D2.2Mpc, 0.09/pixel 1pc) (data retrieved from
the HST Data Archive, Relaño et al. 2006).
4
Derivation of the extinction map

• Extinction maps derived using H? and H? images
  assuming a screen dust distribution (RV3.1)
  (Caplan Deharveng 1986).

• Derive Balmer decrement for regions where S/N gt 3
  both for H? and H?

• Two components to extract the extinction map
• - HII regions T 11000K (change of 1500K
  gives 0.02mag. difference)
• Diffuse ionized gas
• different temperature T6000K (Reynolds Cox,
  1992) (2000K change 0.05mag.difference
• contamination from shock excited gas (an
  overestimate of 0.15mag)

Smoothed Hß image at 6resolution, overlaid with
contours of the Ha emission.HII2, HII4, HII7,
HII5 Most luminous HII regions of the galaxy
catalogued by Waller (1991)
5

• Extinction Good agreement with previous values
  using other observational techniques and
  obtaining only results for some areas (Devost et
  al. 1997, González-Delgado et al. 1997, Origlia
  et al. 2001 and Kobulnicky Skillman 1997).
• Maximum value of internal extinction about
  A(H?) 0.8 mag (1.3 mag Galactic extinction) ?
  in spite of low metallicity locally considerable
  extinction
• Extinction structure shows a pronounced shell
  around SSCA spatially coincidence with the
  expanding shell found by Martin (1998).
• General resemblance to H? map, but
• Clearer shell structure in extinction
• Maximum of extinction lies outward of maximum of
  H?

AHa
Total extinction map (6) with contours of the
extinction overlaid. Contours level 0.9, 1.1,
1.3, 1.5, 1.7, 1.9 and 2.1 magnitudes.
Overlay of extinction map with Ha map of 6
6
Comparison to Spitzer 8 and 24µm data
Color 24µm Contours Ha
Color 8µm Contours Ha
Color 8µm Contours Extinction
Color 24µm Contours Extinction

• General correspondance between extinction/Ha and
  8/24µm, but
• Clear offset between maximum of extinction and
  of 8µm

7
Extinction, H? and emission at 8 and 24µm in NGC
4214
8 µm (color)
Extinction (contour)
24 µm (color)
Extinction (color) and H? (contours)
8

• Comparison of dust emission, extinction and gas
• Dust emission at 850?m (cold dust), CO
  observations (2) follow extinction.

Extinction map with contours of the dust emission
at 850mm at 15 resolution. The asteriscs mark
the position of the GMC detected by Taylor (1999).
An HI ridge surrounds the western and southern
part of the extinction shell.
Extinction map with contours overlaid of the HI
intensity (Stil Israel 2002) at 13.5.
9
NGC 1569

• Comparison of dust emission, extinction and gas
• Dust emission at 850?m (cold dust), CO
  observations (2) follow extinction.
• BUT CAUTION In NGC 4214 this is not the case --gt
  850?m emission more extended

NGC 4214
Extinction map with contours of the dust emission
at 850?m at 15 resolution. The asteriscs mark
the position of the GMC detected by Taylor (1999).
Extinction map with contours of the dust at 850?m
at 15 resolution. Crosses are positions of peaks
of CO (Walter et al. 2001)
10
Position of the cut
Spatial profiles starting at SSCA and going
eastwards at a PA90deg done for A(Ha) and
various wavelength refers to intrinsic extinction
within NGC 1569 after subtracting the Galactic
contribution.

• - Spatial displacements of the maxima of the
  emission profiles.
• Sequence from inwards to outward
• H?, 24 µm ? 8 µm ? (850 µm)/A(Balmer) ? HI
• Offset between H? and A(Balmer) 6
• ? Dust is accumulating in the outer regions of
  the ionized shell surrounding a region where
  ionized gas and dust could be partially mixed.

11
Origin of the extinction distribution

• The extinction shell could be explained by the
  accumulation of dust at the border of the ionized
  shell identified by Martin (1998). Kinematically
  plausible
• Kinematic age of the expanding shell 1Myr,
  consistent with the age of the SSC A (t gt 7Myr,
  Hunter et al. 2000)
• The kinetic energy injected by the stellar winds
  from the SSCA (3x1052erg, Starburst99) is
  compatible with the energy required to move the
  dust mass contained within the shell (3x1051erg)
  .

12
Small scale extinction structure

• Typical size of structures 1 (10pc).
• In NGC 4214 (Hermelo et al. in prep.) and NGC
  604 (Maíz-Apellaínz et al. 2004) variations of a
  similar spatial scale visible.
• Possible, we are seeing inhomogenities in dust
  distribution, possibly due to clumps

• High resolution (0.3) Ha/Hb line ratio map.
• Fluctuations are above values expected from noise
  in H?/H? images

13
Interstellar dust within HII regions24?m
emission as a SF tracer

• A good correlations exists between the luminosity
  at 24?m, L(24?m), and the extinction corrected
  H? luminosity, Lcorr(H?), of star-forming
  regions in different galaxies
• ? L(24?m) emission can be a good tracer of the SF
  (e.g Calzetti et al. 05, 07 Pérez-González et
  al. 06 Kennicutt et al. 07, Wu et al. 05,
  Alonso-Herrero et al. 06 for LIRGs and ULIRGs)
• Deviation from the correlation are attributed to
• Dispersion atributed to uncertainties in
  attenuation estimations (Pérez-González et al.
  06)
• Metallicity effects (Calzetti et al. 2007)
• ? Test the relation with metallicity with
  high-resolution data of dwarf galaxies

14

We derived L24?m and LH?_corr (with matched
apertures) for HII regions. We see The HII
regions in low metallicity dwarf galaxies are
located at the same position in the plot as those
in high metallicity galaxies. No trend with
metallicity is visible.
A possible interpretation The accumulation of
dust around HII regions might produce a
relatively uniform dust opacity around these
regions, largely independent of the metallicity
of the galaxy.
logL(24?m)-7.281.21xlogL(H?corr)
Relaño et al. (2007)
Metallicities M51 (Z8.9), M81 (Z8.7), NGC
6822 (Z8.1) NGC 1705 (Z8.2) NGC 4214
(Z8.2) NGC 1569 (Z8.2)
There might exist a lower threshold for the
accumulation of dust in HII regions in order to
support large SFR ( 10-4-10-1 Msun yr -1).
15

Correlation tested for the total emission of
galaxies
Metallicities (12log(0/H)) High Z Zlt8.0 Med
Z 8.0ltZlt8.5 Low Z Zgt8.5
Relaño et al. (2007)

• Comparison of the correlation with a set of low
  metallicity dwarf galaxies
• Low metallicity galaxies fall below the linear
  fit, whereas the high-metallicity galaxies follow
  it.

L(24?m) vs. extinction-corrected L(Ha) for a
sample of galaxies with different metallicities
A trend of increasing deviation from the
correlation marked by high-metallicity galaxies
with decreasing metallicity is visible
16
Conclusions

• Even in low-metallicity environment the dust
  extinction can locally be high.
• The dust extinction derived with high-resolution
  data from the Balmer decrement shows indications
  of small scale structures, indicative of a clumpy
  dust structure or a complicated geometrical
  distribution
• The dust extinction shows a shell-like
  distribution around large HII regions.
• The maximum of the dust extinction is outside the
  maximum of the Ha emission.
• The 24?m emission integrated over HII regions is
  a good SF indicator, practically independent of
  metallicity. The 24 mu emission, including the
  diffuse emission, does show a dependence of
  metallicity.