Received 13th March 1998
Abstract.
We have revisited the metallicity-luminosity relationship for a homogeneous
sample of dwarf irregular galaxies.
Previous investigation indicated that such a relationship exists in accordance
with a closed-box scenario of galaxy evolution.
In this work new and better distance determination as well as a recalculation
of the metallicities from the previously published emission lines intensities
were obtained.
The conclusion is that no relationship between the oxygen abundance and
the absolute magnitude in the B band is found.
1. Introduction
Many attempts have been made in other to establish a relationship between the
luminosity (L) and the metallicity (Z) of galaxies.
These quantities are normally deduced from the absolute magnitude in some
colour and the oxygen abundance.
Such a relationship (hereafter the Z-L relationship), with the
metallicity proportional to the luminosity of the galaxy, could give some clues
about the evolution of the systems.
For the case of the galaxies on the Hubble sequence, the attempts have been
fruitful with a rather well defined Z-L relationship for ellipticals
(Terlevich et al 1981; Vigroux et al. 1981) and spirals (Vila-Costas and
Edmunds 1992).
The situation for dwarf galaxies is more confusing.
Even if a Z-L relationship has been claimed to exist in dwarf spheroidals
(Aaronson and Mould 1985) and irregulars (Skillman et al. 1989, hereafter SKH;
Richer and McCall 1995, hereafter RM), Vilchez (1995) found some deviations
for dwarf irregulars (dI's).
The Z-L relationship was first established for elliptical galaxies and
was based on a relation between broad-band colours and Mb
(Baum 1959) and between broad-band colours and metallicity (Faber 1973).
In order to check for a relation between metallicity and luminosity for dI's
we reanalysed, critically, published spectra and investigated the methods for
determining the luminosity of the objects.
2. Definitions
In order to reduce the scatter in the Z-L relationship, or at least to
know if it has a physical meaning or just reflects differences in the galaxies,
a homogeneous sample of galaxies should be used.
A definition of a dwarf galaxy was obtained for this purpose.
2.1. Definition of dwarf galaxy
In Fig. 1 is shown the relationship
between the absolute magnitude (Mb),
and the optical radius (r) of a sample of 176 irregular galaxies
selected from the Tully catalog (1988).
A dwarf irregular galaxy was defined such that the
Mb ≥ -17 and r ≤ 3 kpc.
A more thorough discussion regarding this definition can be found in
Hidalgo-Gámez and Olofsson (1998).
We selected all the galaxies fulfilling these conditions closer than
5 Mpc from the Galaxy, which were a total of 45.
[Click here to see Fig. 1!]
3. Determining the key parameters
3.1. The distance
The distance to most of the nearby galaxies is not very accurately known due
to the lack of reliable determination methods.
Whenever possible, Cepheids were used.
This was the case for only ten galaxies.
For the others, a weighted averaged value of the distance were obtained using
the most reliable measurements.
The uncertainties in the Mb was calculated as
Δ Mb = (Δ D / D) 5 log e.
3.2. The metallicity
The metallicities were recalculated from the published emission line
intensities, corrected for internal extinction and underlying stellar
absorption whenever possible.
We used the temperature-sensitive method, which is based on the forbidden
[O III] λ 4363 Å line, for
calculating the metallicity because it is probably less affected by intrinsic
uncertainties when the data are of good quality.
The main problem is that not all the galaxies have spectra in which the
[O III] λ 4363 Å line is measurable
and this condition was fulfilled for only 15 galaxies of the 45 of our sample.
The uncertainties were calculated through the expression
Δ x =
∑i (δ x / δ y) Δ y
where x and y symbolizes the dependent and independent variables,
respectively.
In the best case, an uncertainty of 10% in the oxygen abundance determination
was obtained.
The main objection to this method of calculating the metallicity is
that the nebular forbidden lines necessary for calculating the electronic
temperature only appear in high excitation regions.
The semiempirical methods (e.g. Pagel et al. 1979) are used in absence of
these lines, but the uncertainties in the derived oxygen determination using
these calibrations are probably larger.
4. Results
The revised Z-L diagram is shown in
Fig. 2.
The relationship obtained by SKH as well as a 1 σ deviation from
this is also visualized.
Some aspects need to be considered.
Each point in Fig. 2 represents an
H II region of a galaxy.
In the cases when there are metallicity measurements from several
H II regions all of them are plotted separately.
The oxygen abundance is not averaged over the whole galaxy because much more
information than the published one is needed.
In some cases, large differences in the metallicity between different regions
of a galaxy are found.
This could reflect physical variations in the oxygen abundance or different
excitation and ionization conditions.
The linear correlation coefficient of our sample is
rl = -0.52.
This correlation coefficient was calculated with the averaged oxygen abundance
for the whole galaxy because two global quantities should be compared.
[Click here to see Fig. 2!]
As evident, our results differ strongly for the previous ones.
Actually, 75% of the galaxies in common with SKH's sample fall outside the
1 σ limit.
There could be several reasons for the discrepancies.
The first could be the method for calculating the metallicity.
In order to check if the different methods will give different results
in the Z-L plane we calculated the metallicity with one of the
semiempirical methods (Pagel et al. 1979).
The results is plotted in Fig 3.
No noticeable improvements are seen.
Another problem could be the distance determination.
As mentioned earlier, the distances to these galaxies are quite uncertain.
To investigate this further, only galaxies with reliable distance
determinations, mostly those which have Cepheid light curves, were chosen.
The linear correlation coefficient was rl = -0.20 for 20
H II regions in 10 galaxies.
Another important difference between this work and the previous ones is
the exclusion of more luminous galaxies in this study.
It is clear that the results including brighter and/or larger galaxies are
more in line with the work of SKH and RM, with more luminous galaxies showing
higher metallicities (see Fig. 9 in Hidalgo-Gámez and Olofsson,
1998).
The physical environment could also play a rôle in the sense that
dwarfs with nearby large companion galaxies could have a higher metallicity
possibly due to the higher star formation rate triggered by the interactions.
We also studied this situation, dividing our sample into isolated and
non-isolated galaxies.
No clear distinction is seen between these two categories
(Hidalgo-Gámez and Olofsson 1998).
However, one can question whether the oxygen abundance and the absolute
magnitude in the B-band are the proper indicators to use when studying
a possible relationship between Z and L.
[Click here to see Fig. 3!]
5. Discussion
It is well recognized that both the oxygen abundance and the B-band
magnitude are closely linked to the onset and cessation of star formation
(Olofsson 1995), and they depend strongly on the evolutionary state of the
galaxy.
The abundance of nitrogen in the interstellar medium is much less sensitive
to the star formation history because it is a product of intermediate mass
stars, with life-time of a few billions years.
The same is true for the near-IR magnitude which traces, mostly, the older
stellar population in galaxies.
It is suggested that a relationship between these two quantities, nitrogen
abundances and near-IR magnitudes, could be a better tracer when studying
the Z-L relationship of dI's.
As mentioned earlier, the Z-L relationship is based on (B-V) vs.
Z and (B-V) vs. Mb.
In the case of irregular galaxies no such relationships are found as shown
in Fig. 4, where data from 89 and 15
galaxies, respectively, are plotted.
[Click here to see Fig. 4!]
6. Conclusion
After revisiting the Z-L relationship for the case of dI galaxies
the main conclusion is that, with the present data, no relationship between
the oxygen abundance and the absolute magnitude in the B-band are found.
We are presently obtaining high quality data for a sample of dwarf irregular
galaxies.
These new data will, hopefully, give a more conclusive answer to the existence
or non-existence of a Z-L relationship of dI galaxies.
Acknowledgments.
A.M.H.G. wishes to thank Dr. N. Bergvall for financing the attendance to this
workshop.
We thank L. Cairos, D.I Méndez and K. Noeske for many valuable comments
and discussions, D. Delgado, D.I. Méndez, M. Henkel and U. Klein for
such a nice SBP.
A.M.H.G. is financially supported by NOTSA.
This research made use of the NASA/IPAC Extragalactic Database (NED) which is
operated by the Jet Propulsion Laboratory, California Institute of Technology,
under contract with the National Aeronautics and Space Administration.
References
- Aaronson M., Mould J., 1985, ApJ 290, 191
- Baum W.A., 1959, PASP 71, 106
- Faber S.M., 1973, ApJ 179, 731
- Hidalgo-Gámez A.M., Olofsson K., 1998, A&A 334, 45
- Olofsson K., 1995, A&A 293, 652
- Pagel B.E.J., Edmunds M.G., Blackwell D.E., Chun M.S., Smith G., 1979,
MNRAS 189, 95
- Richer M.G., McCall A.L., 1995, ApJ 445, 642 (RM)
- Skillman E.D., Kennicutt R.C., Hodge P.W., 1989, ApJ 347, 875 (SKH)
- Terlevich R., Davies R.L., Faber S.M., Burstein D., 1981, MNRAS 196, 381
- Tully R.B., 1988, Nearby Galaxies Catalog, Cambridge Univ. Press
- Vigroux L., Chieze J.P., Lazareff B., 1981, A&A 98, 119
- Vila-Costas M.B., Edmunds M.G., 1992, MNRAS 259, 121
- Vilchez J.M., 1995, AJ 110, 1090
Links (back/forward) to:
| First version: | 06th | August, | 1998
|
| Last update: | 08th | October, | 1998
|
Jochen M. Braun &
Tom Richtler
(E-Mail: jbraun|richtler@astro.uni-bonn.de)