next previous
Up: Multicolour deep CCD

11. Field star contamination in the clusters

Stars below tex2html_wrap_inline3420 in NGC 3228; below tex2html_wrap_inline3422 in NGC 4103; below tex2html_wrap_inline3424 in NGC 5662 and below tex2html_wrap_inline3426 in NGC 6087 show a large spread in colour at a given brightness level, compared to that expected from observational errors and differential reddening. The main reasons for such a spread could be the presence of field stars; binaries and peculiar cluster members. The field stars located in the galactic plane, at about the cluster distances and with similar reddenings, will occupy almost the same area in the colour-colour (CC) and CM diagrams as the cluster members. It is therefore difficult to separate cluster members from field stars only on the basis of their closeness to the main populated area of the CM and CC diagrams. However, the possibility of cluster membership is small for stars located well away from the cluster MS in most of the CM and CC diagrams. As the present observations are unable to quantify the field star populations in the vicinity of clusters under study, the statistically expected number of cluster members among the fainter stars can not be determined in the same way as we did in NGC 4755 (Sagar & Cannon 1995). However, the following procedure has been adopted for this purpose. The frequency distributions of stars are derived from the tex2html_wrap_inline3428 CM diagrams (see Table 12 (click here)). These diagrams were preferred over other CM diagrams because of better data completeness in them. Inspection of our CCD frames and an estimate of the numbers of crowded images indicate that we are tex2html_wrap_inline3430 85% complete in general. To avoid the effect of relatively large (tex2html_wrap_inline3432 50%) data incompleteness, the analysis was restricted to the brightness level which is about 1 mag above the limiting magnitude of our observations. To derive the frequency distribution of stars, the tex2html_wrap_inline3434 CM diagram is divided into two magnitude bins, namely brighter and fainter; and into three colour bins called ``below", ``near" and ``above MS". The brighter magnitude bin represents the region in which the cluster main sequences are clearly visible, while in the fainter magnitude bin, any cluster sequence is hardly distinguishable from the field star contamination. Stars lying within two standard deviations in tex2html_wrap_inline3436 either side of the ZAMS are put in the colour bin called ``near MS" and tex2html_wrap_inline3438 denotes number of such stars in Table 12 (click here). The width of the colour bins increases with decreasing brightness. Stars located towards the blue and red sides of this bin are put in the colour bins named ``below MS" and ``above MS" respectively. tex2html_wrap_inline3440 and tex2html_wrap_inline3442 denote the corresponding numbers in Table 12 (click here). This table clearly shows that:

(i)
 in the brighter magnitude bin for a cluster, the number of stars in the ``near MS" bin are significantly larger than either tex2html_wrap_inline3444 or tex2html_wrap_inline3446, indicating that the cluster sequences are generally well defined (see Sect. 8).

  table758
Table 12: Frequency distribution of the stars in the tex2html_wrap_inline3448, tex2html_wrap_inline3450 diagram for the clusters under study derived from our sample. tex2html_wrap_inline3452, tex2html_wrap_inline3454 and tex2html_wrap_inline3456 denote number of stars below, near and above the ZAMS respectively while tex2html_wrap_inline3458 denotes number of cluster members expected in the fainter magnitude bin assuming Salpeter mass function slope

  figure775
Figure 17: The luminosity functions of the ``near MS" (tex2html_wrap_inline3502) stars and the expected cluster members (tex2html_wrap_inline3504) are shown by solid and dotted histograms respectively. N is the number of stars in one magnitude bin of tex2html_wrap_inline3508

  figure781
Figure 18: The tex2html_wrap_inline3510, tex2html_wrap_inline3512 diagrams are plotted again in a), b), c) and d) for NGC 3228, NGC 4103, NGC 5662 and NGC 6087 respectively. Theoretical stellar evolutionary isochrones taken from  Iben (1965) for the pre-MS phase of evolution are plotted. The theoretical stellar birthlines taken from Palla & Stahler (1992) are shown by dotted lines

(ii)
 in the fainter magnitude bin for NGC 3228 and NGC 6087, the numbers of stars below the MS are very large in comparison to the value of tex2html_wrap_inline3514. This may indicate that background field star contamination is very large in the direction of the imaged regions of these clusters. In the case of NGC 4103 and NGC 5662, foreground field star contamination seems to be more dominant. In order to get an idea of the number of cluster members which might be present in tex2html_wrap_inline3516 for the fainter bin, we performed the following simple exercise. Assuming that (a) the mass function of the clusters under discussion can be represented by Salpeter's (1955) mass function (which may be justified in the light of recent studies of young star clusters of our Galaxy (cf. Sagar et al. 1986; Sagar 1993; Phels & Janes 1993) as well as of the Magellanic Clouds (cf. Sagar & Richtler 1991; Sagar 1993 and references therein)) and (b) all the stars near the MS in the brighter bin are cluster members, the number of cluster members expected (tex2html_wrap_inline3518) in the fainter bin are derived. They are listed in the last column of the Table 12 (click here). The luminosity functions of the tex2html_wrap_inline3520 and tex2html_wrap_inline3522 are presented in Fig. 17 (click here) by solid and dotted histograms respectively. Except for NGC 3228, the value of tex2html_wrap_inline3524 is always less than tex2html_wrap_inline3526 indicating clearly that not all MS stars are cluster members. A comparison of these numbers and Fig. 17 (click here) indicate that large field star contamination is present in the case of NGC 5662 and NGC 6087. To know the actual number of cluster members in the tex2html_wrap_inline3528 of the fainter bin, their precise proper motions and/or radial velocities are required. In the absence of such data for the stars observed by us, it is difficult to separate unambiguously the cluster members from the field stars. Nevertheless, there is evidence from visual inspection of several of the cluster CM diagrams that the number of intermediate brightness stars does not increase as rapidly as one might expect, down to the level at which field star contamination makes it impossible to recognise the cluster main sequence. This is particularly true in the case of NGC 6087, where there seems to be a real deficiency of stars for tex2html_wrap_inline3530. The small proportion of faint cluster members, and the low gradient of cluster number density across our relatively small fields, clearly make it impossible to use the spatial distribution to help select cluster members or locate the faint MS.


next previous
Up: Multicolour deep CCD

Copyright by the European Southern Observatory (ESO)
web@ed-phys.fr