Proceedings of the Workshop
"The Magellanic Clouds and Other Dwarf Galaxies"
of the Bonn/Bochum-Graduiertenkolleg

A photometric method to study the Wolf-Rayet content

of compact H II regions in nearby galaxies

P. Royer, J.-M. Vreux, and J. Manfroid

Institut d'Astrophysique, Université de Liège, 5, Avenue de Cointe, B-4000 Liège, Belgium

Received 06th March 1998
Abstract. We present a five-filter photometric system aimed at WR classification. It is shown that this system allows not only the well-known easy separation between the WN and WC spectral types, but also the discrimination of subgroups among the WN and the WC. These subgroups are well related to the classical subtypes. We also show that the system keeps important discriminating power on the spectral types even when the spectra are contaminated by a companion. This photometric classification is particularly interesting to conduct evolutionary studies in crowded fields where spectroscopic follow-up is not possible.

1. Introduction

With a few exceptions (Corso 1975; Armandroff & Massey 1985), the various surveys for Wolf-Rayet (WR) stars conducted so far consist of two-step procedures. The first one is a search for WR candidates by photometry (Schild & Testor 1991; Shara & Moffat 1982; Shara et al. 1991) or by slitless spectroscopy (Testor et al. 1996; Morgan & Good 1985; Westerlund et al. 1983). The second step is a spectroscopic study to precise the types and subtypes of the candidates. However, spatial crowding rules out spectroscopic analysis in the densest regions of clusters and remote galaxies. Clearly, these important fields of research are not compatible with a spectroscopic follow-up.

Using narrow-band filters, several authors achieved a differentiation between WN and WC spectral types and eventually a rough selectivity among the WN stars. It has been shown in Royer, Vreux & Manfroid (1998, hereafter RVM) that an appropriate photometric system can give meaningful discrimination not only between WN and WC stars but also at the level of spectral subtypes. In this paper, we emphasize simulations showing that much more than simple WN/WC separation can be achieved even when real or spurious WR binaries are observed with the proposed set of filters.

2. Filters, data and color diagrams

The main characteristics of the filters are given in RVM, along with the actual choice of color indices used in our color diagrams. Let us just say that they correspond to a normalisation of the flux in the spectral lines by the underlying continuum, which brings the non - WR stars (having no or weak spectral features at those wavelength) to cluster around the origin of the axes.

The color diagrams are constructed through synthetic photometry performed on the spectra of the catalog of Torres-Dodgen & Massey (1988, hereafter TM) for the WR stars and on those from Jacoby, Hunter & Christian (1984, hereafter JHC) for the non - WR stars. Observational data of both WR and non - WR stars are also used. The complete data set gathers 117 WR spectra from TM and 28 non - WR spectra from JHC, 37 observed WR stars and 36 observed non - WR stars. Full details are given in RVM.

3. Influence of a companion

It has been shown in RVM that our photometric system allows a good discrimination not only between non - WR stars, WN stars and WC stars, but also between WNEs, WNEw, WN7 and WN8-9 stars for the WN class and between WCE, WC7-8 and WC9 stars for the WC class.

However, observing crowded regions and remote galaxies will lead us to be confronted with unresolved binary stars, real (physical binaries) or spurious (line of sight coincidence). We simulated such a situation by mixing the spectra of the WR stars with those of non - WR stars before performing the synthetic photometry. Figure 1 shows some examples of such simulations. The curves represent the displacement of the representative points of four WR stars (namely Brey 1, WN3s; WR 35, WN6w; WR 84, WN7w and WR 66, WN8w) when contaminated with an apparently more and more luminous ``companion'' O5V star. The small dots on the curves correspond to a 5% step of contamination, from pure WR star (0%) to pure companion star (100%). The asterisks on the curves indicate 50% of contamination, i.e. WR star and companion having the same apparent magnitude. It must be kept in mind that physical binaries on each of these curves would correspond to different contamination levels depending on the real difference in absolute luminosities between both components.

We performed simulations with a wide variety of spectral types for both components. Nevertheless, the spectral type of the non - WR star is of minor importance because the displacement of the representative points of the WR stars in our color diagrams comes from the ``dilution'' of the light in the spectral lines, not from the spectral characteristics of the companion.

[Click here to see Fig. 1!]

4. Conclusions

An examination of Fig. 1 as well as various simulations not shown here indicate that not only the WN/WC distinction but also the separation between the WN subtypes is basically maintained even when the light of the WR star is diluted by that of one or sometimes several physical companions, the only exception being the low-luminosity early subtypes.

Combined with state of the art techniques to deconvolve crowded fields, the photometric system discussed here should allow to get information on the upper part of the IMF in the cores of H II regions, well beyond the limit reachable through the usual spectroscopic methods.

Acknowledgments. The authors are indebted to the Fonds National de la Recherche Scientifique (Belgium) for multiple supports. This research is also supported in part by contract ARC 94/99-178 ``Action de recherche concertée de la Communauté Française'' (Belgium). Partial support through the PRODEX XMM-OM Project is also gratefully acknowledged.

References

Links (back/forward) to:
First version: 25thJuly,1998
Last update: 30thAugust,1998
Jochen M. Braun   &   Tom Richtler
 (E-Mail: jbraun|richtler@astro.uni-bonn.de)