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

3. Stellar density in and around the clusters

In order to investigate the stellar surface density distribution in and around the clusters we subdivided the entire area into 52 square cells with 20 pixels (corresponding to ) length each. We counted the number of stars found within each cell. In Fig. 2 (left) one cell corresponds to one pixel of the star density plot. The counting was carried out on the ALLSTAR-output tables, which contain the coordinates of each measured star. We see an enhanced star density between the cluster pair. The apparent stellar bridge between the components contains as many as 25% of the stars counted in each of the clusters. To make density structures and thus the stellar bridge between the two clusters better visible we applied a average filter for image smoothing (see Fig. 2, right). Since we expect especially high values in the stellar bridge compared to the surrounding field, a median filter - which does not consider very high or very low pixel values - seems not to be appropriate for our purpose. However, we also tried a median filter and found no differences concerning the features of the resulting images.

Fig. 2. Left: Investigating the stellar surface density distribution we see enhanced stellar density between the two components while no enhancement between the binary cluster candidate and KMHK 1019 can be seen. We plotted 8 grey values, reaching from black (no star per cell) to white (7 stars per cell). Right: To make density structures and thus the stellar bridge between the two clusters better visible we applied a average filter for image smoothing. Note that one cannot use this image for star counting as the pixel values are averaged to repress noise

No signs of an increased star density reaching from KMHK 1019 towards the binary cluster candidate can be seen.

We investigated the statistical probability of seeing such a stellar bridge due to chance density fluctuations. For this purpose we removed all stars outside a 60 pixels radius around the components of the cluster pair and carried out 50 artificial star experiments with ADDSTAR running under DAOPHOT II. Each time we added the same number of stars at random coordinates as had been removed initially. Star density plots were created in the same way as described above. We found 12 plots out of 50 showing larger star density somewhere between - and connecting - the star clusters. In most cases the stellar bridge connecting the components is very thin. Only three plots showed stellar bridges wider than one cell. In Fig. 3 three examples of our artificial star density plots are shown: no bridge, small bridge, and wide bridge. Our artificial star experiments show that in one out of four cases a stellar bridge may occur due to statistical density fluctuations in the field. Thus, the probability of a stellar bridge which is not due to statistical star density fluctuations does not reach a confidence level of 95 %.

Fig. 3. Star density plots based on artificial star experiments. Left: no bridge occurred between the star clusters, middle: a small bridge connects the components of the cluster pair, right: the stellar bridge is more pronounced. This last case occurred only three times out of 50 star experiments and shows that there is a low likelihood of a pronounced stellar bridge produced by random fluctuations (6%)

© European Southern Observatory (ESO) 1998

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