Diapositive 1

1
Is the ISM in Starburst galaxies well mixed?
presented by Vianney Lebouteiller (IAP) coll.
with Daniel Kunth
2
Team / Collaborators

• Thesis of Vianney Lebouteiller
• Team of the Institut dAstrophysique de Paris
  (IAP)
• french team (A. Lecavelier des
  Etangs, J.-M. Désert, G. Hébrard, R. Ferlet, A.
  Vidal-Madjar)
• LERMA - Obs. de Paris J. Lequeux
• JHU Baltimore A. Aloisi

3
Outline

• Problematic self-enrichment of the HII regions
  ?
• A new approach study of the neutral gas
• FUSE
• method and analysis
• Results on Blue Compact Dwarfs
• The next step Giant HII regions
• the example of NGC 604
• Results on NGC 604
• neutral gas homogeneity
• abundances ratio
• modelling
• Conclusions Perspectives

4
Origin of the problem
IZw18

• The problem of mixing in the ISM is a genenal one
• Moreover, Kunth Sargent (1986) pointed out
• the lack of extremely metal-poor galaxies in
• emission-line surveys hence proposed
• Self-enrichment hypothesis
• Selfpollution by fresh released metals in situ
  by SNe,
• providing rapid cooling of the hot gas during
  the lifetime of the starburst.
• Pristine gas in the outer regions ?

5
Evidences against/for abundances discontinuities

• Small scale abundances inhomogenities suspected
  in a few objects
• NGC5253 (Walsh Roy, 1989 Kobulnicky et al.
  1997)
• local pollution in IIZw40 (Walsh Roy, 1993)
• and Mkn 996 (Thuan et al. 1996) N/H
  overabundances
• ? attributed to WR winds. But not true in all
  young starburts (Kobulnicky, 1999 Oey Schield,
  2000)
• Abundances are similar to those of young stars

IZw18

• IZw18, SBS0035-052,
• SMC 6 regions log O/H8.13 0.08
• LMC 4 regions log O/H8.37 0.25
• (Russel Dopita 1990)
• a-elements in hot gas is often over
• solar e.g. in the LMC (Dennerl et al.
• 2001) ? Delayed mixing ?

6
Two models
Recchi et al. 2002
Tenorio-Tagle 1996
ejecta and winds carry out metals but gas
immediately cools down and enrich the ISM
SNe expell the gas, creating a superbubble
evolving in a low density spherical halo
disperse and mix
few 109 yrs
few 106 yrs
Mixing timescale
discontinuities
no discontinuities
7
FUSE

• 900-1200 Å
• HI, OI, NI, FeII,
• SiII, ArI, ...

8
Studying the HI metal composition
9
Abundances determination

• Profile fitting procedure Owens (M. Lemoine)
• returns most likely values (N, b, T, ...)
• Owens derives an estimation of the errors
• (including uncertainties on the LSF, the
• continuum,...)
• One homogeneous cloud
• Deriving abundances from the column densities
• ionization structure assumed a priori

Fig Example of fitted lines
H N O Si P Ar Fe
H I (13.60 eV) N I (14.53 eV) O I
(13.62 eV) SiII (16.34 eV) P II (19.72 eV) Ar
I (15.76 eV) FeII (16.18 eV)
10
Results Blue Compact Dwarfs
HI region (UV abs. lines) HII region (opt. em.
lines)
log (X/H)
log (X/H)
Refs Lebouteiller et al. (2003) for IZw36,
Lecavelier et al. (2003) and Aloisi et al. (
2003) for IZw18 Thuan et al. (in preparation)
for SBS 0335-052, Thuan et al. (2002) for
Markarian 59
(NGC1705 poster n3 A. Aloisi and NGC253 poster
n14 J. Cannon)
11
? ? ?
12
? Observations of individual HII regions in
spiral galaxies
M33

• Relatively high S/N, sightlines less complex
• Avoid possible systematic errors, study the
  effects of the multiple sightlines, ...
• Unique possibility to model, to know the
  ionization stucture, etc...

1
13
The case of NGC604

• 3-5 Myrs (Hunter et al. 1996),
• 4-5 (Wilson Matthews 1995)
• 200 pc HII region size

14
NGC 604 spectrum
CIII Pcygni feature
OVI Pcygni feature
Metals from M33 HI from M33
15
HI column density determination
Profiles before convolution Profiles after
convolution Red total profile

• log N(HI) 21.010.05

16
HI withSTIS
20.73 0.15 star O4 Ia, V17.0
20.86 0.18 star O4 Iab, V18.2
Lyman a
20.52 0.19 star O4, V19.2
23/18
(fig. of Bruhweiler at al. 2001)
17
Mimic FUSE data

• Construction of the global spectrum mean,
  weighted by the stars magnitudes

• N(HI)20.96

• Do we overestimate HI column density ?

18
Results abundances X/H

• Consistent determinations
• Global underabundance in the neutral gas, even
  for Fe
• Less enriched gas in the sightlines vs.
  enrichment of the HII gas

log (X/H)
X/H
19
Results abundances X/Y

• N/O same as HII region ? N primary, no ionization
  correction
• Ar/O ? little ionization correction may be needed

log (X/Y)
X/Y
20
Modelling of NGC604
PDR
HII region
HI region gt
Coupling with HI
21
Conclusions perspectives

• Difficult interpretation for BCDs needs to be
  validated
• ? Giant HII Regions
• Evidence for pockets of metal deficient neutral
  gas in NGC604
• less chemically evolved gas in the sight lines ?
• are dustier regions invisible in far-UV more
  metallic ?
• Is the neutral gas associated to the HII region ?
• v(HI) always ? v(HII)
• Future work on other giant HII regions in M33
• Investigate hidden saturated lines (...resolution
  effects)

See poster n41 for more details
22
end of the presentation
23
Column densities
24

• NGC 604 MDRS and LWRS observations
• ? test

LWRS
MDRS
HST/ STIS
The derived broadenings are consistent with the
size of the cluster
15
Total broadening/3 (px)
4
? (Å)
25
HI stellar contamination ?
Stellar vs. interstellar HI ? O stars dominate ?
No significant stellar contamination is expected
for Lyß in hot stars stellar population (Robert
et al. 2003)
Observations Synth Mod Spec Theo model
26
NGC 604, some facts
27
(No Transcript)
28
Stellar contamination evidences
Synthetic model spectra (Robert et al. 2003) 0.5
Z? IMF-2.35 Instantaneous burst Uses a stellar
library from observations
Synthetic model spectra (Starburst 99) 0.5
Z? IMF-2.35 Instantaneous burst Uses a stellar
library from observations
C III
29
Modelling of NGC604
HII region
PDR

• Cd fit LWRS independent compare with HIIR cd
• To come PDR modelling

NEUTRAL REGION