Astron. Astrophys. 339, 773-781 (1998)

4. Deriving ages for the star clusters

We derived ages of the clusters SL 538, NGC 2006, KMHK 1019, and the surrounding field star populations by comparing our CMDs with isochrones. The isochrones we used are based on the stellar models of the Geneva group (Schaerer et al. 1993).

4.1. Age determination

We derived CMDs of each cluster by cutting out a circular area with a radius of 100 pixels, corresponding to or 8.3 pc, centered on the optical centre of each cluster. To derive the CMDs of KMHK 1019, which is the smallest of the three clusters, we adopted a smaller radius of 80 pixels, corresponding to or 6.6 pc. An estimation by eye based on the star density plot (Fig. 2) suggests that no or almost no cluster stars are outside this area.

All CMDs are plotted in Figs. 4 to 7. Each cluster CMD has a wide blue main sequence and contains very few supergiants. The width of the main sequence is caused in part by photometric errors, crowding (seeing ) and the presence of Be stars (see Sect. 5). The scarcity of supergiants is well within the expected fluctuations for compact clusters with few stars.

Fig. 4. Colour-magnitude diagram of the star cluster SL 538. We adopt an age of Myr. The derived age is supported by the best-fitting isochrones in all diagrams. Be star candidates (Sect. 5) are marked with crosses

Fig. 5. Same as above, but for the cluster NGC 2006. The adopted age is Myr, which is confirmed by the isochrone fits in all colours

Fig. 6. CMD of KMHK 1019, the smallest of the three star clusters and with the lowest number of stars. The resulting age for the cluster from the best fitting isochrone is 16 Myr

Fig. 7. CMD of the surrounding field. This diagram comprises the mixture of ages of the various field populations: the blue main sequence and the supergiants represent the younger populations, the intermediate age populations show up through the RGB and the pronounced clump. Note the widening of the main sequence due to Be stars (see Sect. 5). Stars with are plotted as smaller dots to keep the isochrones recognizable

Overplotted on the CMDs are the best fitting isochrones. We fitted the isochrones such that the supergiants rather match the blue loops than the quickly traversed subgiant branch. A distance modulus of 18.5 mag (Westerlund 1997) was adopted. The metallicity of the young field population of the LMC was found to be  dex by various authors (Russell & Bessell 1989, Luck & Lambert 1992, Russell & Dopita 1992, Thévenin & Jasniewicz 1992). We therefore adopted Geneva isochrones with Z=0.008 which corresponds to  dex.

Galactic field stars contaminate our observed area. Ratnatunga & Bahcall (1985) estimate the number of foreground stars towards the LMC, and in Table 2 we present their counts scaled to our total field of view ().

Table 2. Number of foreground stars towards the LMC calculated from the data of Ratnatunga & Bahcall (1985), scaled to our field of view of

SL 538 : We see two red and one blue supergiant in the V, CMD of SL 538. The 16 Myr (solid line) isochrone fits all three supergiants well but also the 20 Myr (dotted line) isochrone fits the red supergiants very well. We adopt an age of Myr. The same age is found from isochrone fits to the V, CMD (Fig. 4). All isochrones are based on a reddening of  mag.

NGC 2006 : Four supergiants are located in this cluster, covering a colour range from to 1.5 mag. Both CMDs (Fig. 5) are fit quite well by isochrones with ages of 20 Myr (solid line) and 25 Myr (dotted line). We adopt an age of 22.5 Myr ( Myr) and a reddening of  mag.

KMHK 1019 : This cluster by far is the smallest one with the lowest number of stars. Few data points are located in the red clump and the red giant branch (RGB) region, and it is very likely that these stars belong to an intermediate-age field star population, while the supergiants, on which our age determination mainly relies, are located in the cluster centre and thus we assume that they belong to the star cluster. The main sequence is sparse, especially in the upper part brighter than mag. The best fitting isochrones in both CMDs result in an age of 16 Myr (Fig. 6). The reddening of  mag is the same for all fits.

The surrounding field : The field population comprises a mixture of ages. Apart from a blue main sequence and blue and red supergiants, which represent the young field populations, the intermediate-age field population of the LMC shows up through red giants and the pronounced red horizontal branch clump. We are not able to distinguish between distinct young populations, but the plotted isochrones represent ages which are supported by corresponding supergiants.

The brightest blue supergiants and some of the brightest red supergiants are represented by the 16 Myr isochrone (solid line). Also the 25 Myr (dotted) and 32 Myr (solid) isochrones are supported by bright blue, yellow and red supergiants.

Several supergiants are traced by the 80 Myr isochrone. Note that the redder main sequence stars at  mag are candidate Be stars (see Sect. 5). One could easily mistake them for more evolved stars marking an additional field population with an age of approximately 100 Myr. The stellar density seems to be lower between 80 Myr and 200 Myr which indicates a possible decrease in the field star formation rate.

Along the 200 Myr isochrone and below, the star density is increased along the subgiant branch, again indicating enhanced star formation.

mag is a lower limit to the reddening of the field star populations and corresponds to the blue envelope of the main sequence(s).

The youngest field population is part of LH 77. Our derived age of approximately 16 Myr is in good agreement with the findings of Braun et al. (1997).

4.2. Comparison to earlier photometry

Our study is the first age determination of SL 538 and NGC 2006 that is based on CMDs. Previous studies derived ages based on surface photometry, using different aperture sizes. Age determinations based on integrated colours are less precise than age determinations based on CMDs: Geisler et al. (1997) investigated the influence of a few bright stars on the integrated light of intermediate age star clusters. They conclude that fluctuations in the number of bright main sequence stars and red giants lead to shifts in the integrated colours, which affect the age determination. In Table 3 we present a comparison to the integrated photometry of Bica et al. (1996), Bhatia (1992) and our results.

Table 3. Comparison of earlier age determination based on integrated colours and ours based on CMDs

Though several authors using surface photometry state that integrated colours and thus the derived ages are largely independent from the aperture radius, Bica et al. (1996) and Bhatia (1992) derived quite different ages for SL 538 and NGC 2006: Bica et al. (1996) used an aperture size of and found NGC 2006 to be the older component of the cluster pair. This is in agreement with our results. In contrast, Bhatia (1992) used an aperture radius of and found SL 538 to be slightly older than NGC 2006. Our CMDs exclude ages as young as suggested by Bhatia's (1992) aperture photometry. Bica et al.'s (1996) young age of only 0-10 Myr for SL 538 again is clearly excluded by our data, while the wide age range of 10-30 Myr for NGC 2006 includes our result of Myr. Note that Bica et al.'s (1996) apertures are so large that stars belonging to the neighbouring cluster are also included in their measurement.

© European Southern Observatory (ESO) 1998

Online publication: October 22, 1998
helpdesk.link@springer.de