Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 346, 459-464 (1999)

Previous Section Next Section Title Page Table of Contents

2. Number counts of Be stars in clusters

2.1. Cluster data

Since the fraction of Be stars is rapidly changing with cluster ages (Mermilliod 1982), we must carefully separate ages and metallicity effects. We select clusters in the age range of [FORMULA] age = 7.00 to 7.40 which have their turnoff around the maximum of the Be star distribution. The comparison of clusters of various ages in different galaxies may lead to confusion between age and metallicity effects.

For the Milky Way the clusters were selected from the cluster data base WEBDA (Mermilliod 1999). This data base contains homogeneous photometric data with consistent estimates of reddening, distance moduli and ages. WEBDA also provides indications on known binaries, stellar peculiarities and Be star identifications based on standard spectroscopic criteria. The data base includes the recent survey of stars with H[FORMULA] emission by Kohoutek & Wehmeyer (1997). Some 11 well studied clusters of the Galaxy are in the age interval considered (Table 1). On the basis of their galactocentric distances, the clusters are separated in two groups: one outside and one inside the solar location.


Table 1. Cluster data

For the LMC the photometric data, the age determinations and the identifications of Be stars come from Grebel (1997), Dieball & Grebel (1998), Keller et al. (1999), and Grebel & Chu (in preparation). A distance modulus of 18.50 is taken in agreement with current determinations (cf. Cole 1998). Nine LMC clusters (listed in Table 1) are in the considered age interval. For the SMC the typical cluster NGC 330 is analysed with the data from Grebel et al. (1996). Unfortunately there is no other cluster with reliable Be star identification in or even near the considered age range. The cluster NGC 346 of the SMC, which contains a few Be stars, is much too young and has a turnoff far out of the age range where the maximum of Be stars occurs. Kudritzki et al. (1989) give an age of [FORMULA] yr for NGC 346, which is consistent with the presence of stars up to about 100 [FORMULA] (Massey et al. 1989). This is in agreement with other authors (Cassatella et al. 1996; Haser et al. 1998) who notice the presence of O3 stars in NGC 346.

2.2. Number counts of Be stars in luminosity intervals

We choose to proceed to number counts in given intervals of magnitude [FORMULA] rather than in given ranges of spectral type and [FORMULA] since spectral types are not available for all clusters. Also, we noticed that high rotation strongly modifies the average [FORMULA] and very little the average stellar luminosities (Maeder & Peytremann 1970). Furthermore, massive stars of the same mass, but different metallicites, have rather large differences in [FORMULA], while this is not the case for the luminosities (Schaller et al. 1992). In order to test the independence of the results with respect to the chosen [FORMULA] range we have done number counts in three intervals of [FORMULA] centered on the domain O9 to B3, i.e. [FORMULA] = -4, -2; [FORMULA] = -5, -2; [FORMULA] = -5, -1.4 (the limit [FORMULA] = -1.4 corresponds to the spectral type B3 according to the calibration of Zorec & Briot (1991) for main sequence B-type stars). Fig. 1 shows an example of the HR diagram used for the number counts in the case of NGC 884 ([FORMULA] Persei cluster) with the various magnitude intervals considered.

[FIGURE] Fig. 1. Example of colour-magnitude diagram used for the number counts of Be and B stars in the case of NGC 884 from the WEBDA data base (Mermilliod 1999). The limits of the various magnitude intervals are shown: [FORMULA] = -4, -2; [FORMULA] = -5, -2; [FORMULA]. Be stars are represented by black dots. The isochrone corresponding to log age = 7.1 is represented by a continuous line, and the top of the binary sequence upwards shifted by 0.75 mag. is shown by a broken line.

Table 1 shows the number counts for the 19 clusters in the four different locations considered. The 2nd column gives the [FORMULA] of the age, the 3rd column gives the distance moduli in the SMC and LMC, the galactocentric distance for galactic clusters is given in the 3rd column as well; the following columns give the numbers of B plus Be stars, and those of Be stars in the indicated magnitude intervals. The reference for the data is in the last column.
The average number ratios [FORMULA] stars in the four zones and for the different magnitude intervals considered are given in Table 2. We clearly notice a systematic trend for smaller [FORMULA] number ratios in the sequence of the four groups considered from SMC to LMC to the galactic exterior and interior. The total difference in the [FORMULA] ratios amounts to a factor of about three, which is important and also systematic in relation with the metallicities as discussed below.


Table 2. Number ratios of Be to B+Be stars in the Magellanic Clouds and Milky Way.

2.3. Possible relations with Z

We examine the possible relation between the fraction of Be stars and the local metallicity Z in the regions where these stars were formed. Let us notice that the suggestion to determine whether the frequency of Be stars is somehow related to the metal abundance was also made by Grebel et al. (1992) and Mazzali et al. (1996).

We use the recent data on the metallicity Z for the various zones. For the solar metallicity we take a value of [FORMULA] = 0.018, which is consistent with recent solar models and heliosismological data (Brun et al. 1998). For the chemical gradient in the Galaxy the current values range between [FORMULA] (cf. Shaver et al. 1983; Gummersbach et al. 1998) and -0.05 (cf. Vilchez & Esteban 1996). The group of clusters towards the anticenter has an average galactocentric distance [FORMULA] of about 1.5 kpc larger than that of the Sun (taken to be [FORMULA] kpc), while this average difference is 0.6 kpc for the group of clusters towards the galactic interior. Thus we consider that the average local metallicities for the exterior and interior groups are respectively Z = 0.014 and Z = 0.020 for a galactic gradient of -0.07. For a gradient of -0.05 the Z-values would be 0.015 and 0.019 respectively, i.e. not very different from the latter values.

For the young population in the LMC the range of estimated metallicities is rather large. For example, Olszewski et al. (1991) give [Fe/H] values in the range -0.3 to - 0.42, which corresponds to Z = 0.010 to 0.008. Jasniewicz and Thévenin (1994) find values from [Fe/H] = -0.4 for NGC 1818 to -0.55 for NGC 2004, which correspond to respectively Z = 0.008 and 0.006. Bica et al. (1998) obtained values Z = 0.004 to 0.005. Luck et al. (1998) find a mean [Fe/H] = [FORMULA] for LMC field Cepheids. In this context, an average value of 0.007 seems appropriate. For the SMC field Reiterman et al. (1990), Spite et al. (1991), Grebel & Richtler (1992) give values corresponding to Z = 0.001 to 0.003. For NGC 330, Hilker et al. (1995) give a value Z = 0.002, and the results of Hill (1997) correspond to 0.003 to 0.004 and those of Oliva & Origlia (1998) to 0.001. Luck et al. (1998) find a value corresponding to Z=0.004. Gonzalez & Wallerstein (1999) find Z = 0.002. In this context an average value Z = 0.002 for the SMC is appropriate.

Fig. 2 shows the relation between the fraction [FORMULA] and the local average metallicity. The trend is quite clear for the various magnitude intervals considered. For the sample of the 4 locations there seems to be a clear decrease of the fraction of the relative number of Be stars with the local average initial metallicity.

[FIGURE] Fig. 2. Relation between the number ratio Be/(B+Be) and the local metallicity for the 4 groups of clusters considered in Table 2. To test the validity of the results, the number counts were made in different magnitude intervals, the dots refer to counts made in the magnitude interval [FORMULA] = -5, -1.4, the crosses to the interval -5, -2 and the triangles to the interval -4, -2.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: May 21, 1999