6 Comparison of cluster and field Be star populations

The Magellanic Clouds offer us an opportunity to examine the field population of Be stars in different localities in an unbiased way - a task which is virtually impossible from within our own galaxy due to depth and reddening effects. For the purposes of our survey we can regard each CCD image as essentially a volume and magnitude limited sample of the population.

We have defined the cluster extent by a direct examination of the radial variation in surface density of bright blue stars (V < 17 and V-I < 0.6), as described in Sect. 3. Some stars co-eval with the cluster stars probably exist outside our adopted radii. In particular, we note that Elson et al. (1987) have examined the surface density of stars around a sample of young LMC clusters and suggested that some of these clusters are tidally unbound with up to 50% of their total mass in an unbound halo (although the halo is of very low surface density compared to the background).

The CMDs for both the cluster and field of NGC 330, NGC 346, NGC 1948, NGC 2004 and NGC 2100 are very similar (see Figs. 1b to 6b). This indicates that the field populations contain stars of age similar to that of the cluster. We have investigated the fraction of stars along the main-sequence that are Be stars by binning both emission-line and non-emission-line stars with V-I < 0.5 into 1/2 magnitude bins in V. Figures 8 to 13 show our results. In Table 2, we give the numbers of Be stars and main-sequence stars with V < 17 for each cluster and its surrounding field: we also give the ratio of Be to main-sequence stars (with a 1-$\sigma$ error).

Here we have made the assumption that comparison of the number of Be stars to B stars within a certain luminosity range is a valid comparison. This assumption is challenged by the study of Zorec & Briot (1997) who claim that the location of Be stars within the CMD is the result of a visual luminosity excess. The V excess described by Zorec & Briot rises from 0.0 at B9 to -0.5 at B0 in a linear manner. Here we do not adjust for V excess as we consider the size of the associated uncertainties in the correction similar to the size of the correction itself.

  

Figure 8: a) Histograms of the number of Be stars (shaded) and main-sequence stars (V-I < 0.5, including Be stars) in half-magnitude bins down the main-sequence of NGC 330. b) Same as a) but for the field around NGC 330. c) The ratio of Be stars to main-sequence stars in half-magnitude bins down the main-sequence of NGC 330. d) Same as c but for the field around NGC 330

  

Figure 9: Same as Fig. 8 but for NGC 346

  

Figure 10: Same as Fig. 8 but for NGC 1818

  

Figure 11: Same as Fig. 8 but for NGC 1948

  

Figure 12: Same as Fig. 8 but for NGC 2004

  

Figure 13: Same as Fig. 8 but for NGC 2100

The fraction of Be stars in the field populations lies in the range 0.10 to 0.27. The mean value is a similar to the fraction of Be stars in the Galactic field which is 0.17 (Zorec & Briot 1997). The fractions of Be stars in the clusters is quite variable. The clusters NGC 330 (see also Grebel et al. 1992) and NGC 2100 are particularly rich in Be stars, NGC 1818 has a moderate richness, while NGC 346, NGC 1948 and NGC 2004 have low Be star fractions. In the Galaxy, Zorec & Briot (1997) noted that the Be star fraction reached a maximum of 0.34 at spectral type B1. The fraction of Be stars in NGC 330 (0.34 $\pm$ 0.08) is perfectly consistent with this value. In a forthcoming paper, the correlation of Be star ratio with spectral type will be examined.

  

Table 2: Be star content of clusters and fields

  

Figure 14: The finding chart (R image) for the Be stars around NGC 330. The full field is 10$\hbox{$^\prime$}\times$10$\hbox{$^\prime$}$ while the inner rectangle is $7\hbox{$.\mkern-4mu^\prime$}3\times7\hbox{$.\mkern-4mu^\prime$}9$ and is the part of the field on which the search for stars with H$\alpha$ was performed. North is to the right and east is at the top

In order to compare the Be star fraction near the main-sequence turnoff with the Be fraction for less evolved stars, we have divided the stars into two groups: the brighter stars with V < 16 and the fainter stars with 17 < V < 16, all stars having V-I < 0.6. The Be star fractions for these groups are given for the clusters and their surrounding fields in Table 2. From these ratios, and looking at Figs. 8 to 13, it seems that for most of the clusters and fields there is no statistically significant difference between the Be star fractions near the main-sequence turnoff and further down the main-sequence. There are a few distinct exceptions to this rule: the clusters NGC 330 and NGC 2004 show a significantly higher fractions of Be stars near the main-sequence turnoff than at fainter magnitudes, as does the field of NGC 1818 (where the four brightest main-sequence stars are all Be stars).

  

Figure 15: Same as Fig. 14 but for stars near the core of NGC 330

  

Figure 16: The finding chart for the Be stars around NGC 346. The field size is 10$\hbox{$^\prime$}$ square and the orientation is north to the right, east at the top