Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 327, 562-568 (1997)

Previous Section Next Section Title Page Table of Contents

5. Properties of cluster candidates

We defined as cluster candidates those objects that are clearly unresolved in our V image, i.e. those objects that have FWHM of the PSF smaller than [FORMULA] (see the dots in Fig. 2). The HST image shows that most of these objects are also unresolved at its higher resolution, as one would expect, since even the largest Milky Way globular clusters would appear hardly resolved at Fornax distance. (Note that the fainter clusters and objects in the planetary camera image could not be measured due to the short exposure time of the HST image). We counted all objects within [FORMULA] radius from the geometrical center within the lowest isophotes. These objects are uniformly distributed over the whole area, contrary to the light of the main stellar body and the OB association (see Fig. 4).

Applying the same selection criteria as for the NGC 1427A cluster candidates we counted objects in a background control field [FORMULA] north of NGC 1427A. We subtracted the area corrected background counts from the galaxy object counts in color bins of 0.15 mag as well as in magnitude bins of 0.5 mag Figs. 5, 6 (upper panels), and 7 show the resulting histograms. The color histograms are selected for objects brighter than [FORMULA] mag. As mentioned above not all objects could be measured in the B image. Thus, the [FORMULA] color distribution only contains about 80% of the objects of the [FORMULA] color distribution.

[FIGURE] Fig. 5. The hashed histogram (upper panel) shows the [FORMULA] color distribution of clusters in NGC 1427A after a statistical subtraction of background objects (dotted histogram). The mean error is indicated. Note that the colors of the bluest clusters are contaminated by emission lines of nearby H II regions. The lower panel illustrates the color evolution of clusters as function of metallicity according to Fritze v. Alvensleben & Kurth (1997). The peak color [FORMULA] = 0.7 mag corresponds to an age of about 2.0 Gyr assuming an LMC metallicity.

[FIGURE] Fig. 6. The hashed histogram shows the [FORMULA] color distribution of clusters in NGC 1427A after a statistical subtraction of background objects (dotted histogram). As in Fig. 5 the lower panel illustrates the comparison to color evolution models of clusters.

[FIGURE] Fig. 7. The hashed histogram shows the V luminosity function of clusters in NGC 1427A after a statistical subtraction of background objects (dotted histogram) from the raw counts (dashed histogram). The brightest objects have absolute magnitudes of about [FORMULA] mag comparable to the luminosity of the brightest LMC clusters. The thick line represents the counts of the star clusters in the LMC (Bica et al. 1996) corrected for the Fornax distance and normalized to the luminosity of NGC 1427A.

The [FORMULA] color histogram shows a main concentration of objects around [FORMULA] mag with a tail to bluer colors and a few objects with [FORMULA] mag. Colors bluer than [FORMULA] mag can not be explained by a low metallicity alone, since even the metal poorest clusters in our Milky Way have redder [FORMULA] colors (e.g. Mc Master catalogue, Harris 1996). Concerning the [FORMULA] color most cluster candidates are distributed between [FORMULA] mag. The mean photometric errors are [FORMULA] mag and [FORMULA] mag.

We can estimate age limits for the objects in dependence of metallicity. To do so, we used the models of Fritze v. Alvensleben & Kurth (1997, private communication, update of Fritze v. Alvensleben & Burkert 1995). The lower panels in Figs. 5 and 6 illustrate the dependence between color, age and metallicity for a stellar population that experienced a single star burst like star clusters. In the metal poorest ([FORMULA]) models, the bulk of the clusters with [FORMULA] mag (or [FORMULA] mag) would be younger than 7 Gyr. Only the clusters at [FORMULA] mag (or [FORMULA] mag) could represent old metal poor globular clusters. Assuming a LMC metallicity ([FORMULA]), the reddest clusters ([FORMULA] mag) would have ages of about 5 Gyr, whereas the peak color [FORMULA] mag would be consistent with an age of [FORMULA] Gyr. This age is also consistent with the results in the [FORMULA] color (see Fig. 6). In this case, the colors as well as the ages of the NGC 1427A clusters are comparable to (but perhaps slightly redder/older than) the intermediate age clusters in the LMC (Arimoto & Bica 1989).

The luminosity function of the cluster candidates appears to be complete down to [FORMULA] mag which corresponds to an absolute magnitude of [FORMULA] mag. The brightest objects have absolute magnitudes of the order of the brightest LMC clusters, as for example NGC 1850 ([FORMULA] mag, Bica et al. 1996). The number of clusters in NGC 1427A down to [FORMULA] mag is about 0.75 times the cluster counts in the LMC (see Table 2). Normalized to the absolute luminosity of their host galaxies, the counts are comparable. Fig. 7 shows the luminosity function of the NGC 1427A clusters in comparison to the one of the LMC clusters after correction for the Fornax distance and normalization to the luminosity of NGC 1427A. The LMC cluster counts are based on the list of Bica et al. (1996). We included only objects that are classified as "star cluster". Our analysis shows that NGC 1427A has the richest system observed in an irregular galaxy more distant than the Magellanic Clouds up to now. The old cluster population of the LMC has colors of about [FORMULA] mag and absolute magnitudes [FORMULA] mag, which is close to our completeness limit. The distribution of the old LMC clusters is very extended around the LMC, which implies that we will miss corresponding objects in NGC 1427A as a consequence of our statistical background correction. Both effects result in a bias against the detection of true old clusters in our data.


Table 2. Cluster counts in NGC 1427A and the LMC

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: April 6, 1998