Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 358, 910-922 (2000)

Previous Section Next Section Title Page Table of Contents


[FORMULA]Cen is a very massive globular cluster, with a relatively low central density. Hertz & Grindlay (1983) reported five sources A-E near [FORMULA]Cen. Sources A, D and E all are well outside the cluster core, and it appears that only source C is clearly related to the globular cluster (e.g. Verbunt et al. 1995). Source E has tentatively been identified with a foreground K star (Margon & Bolte 1987). The sources A and D have been identified with foreground M stars, on the basis of better positions for the X-ray sources obtained with Einstein and ROSAT HRI observations (Cool et al. 1995a). A ROSAT PSPC pointing indicates that source C, near the cluster center, is composed of two sources of comparable luminosity (Johnston et al. 1994).

We have analysed all observations listed in Table 1. The 1992 and 1993 data have been reported on before by Cool et al. (1995a). As expected, we detect the largest number of sources in the longest observation, that of July 1995. We use this observation as the basis for comparison with the other observations.

3.1. Source list and membership probability

In the July 1995 observation we detect twelve sources, listed in Table 2. Eight of these sources have been detected before with Einstein or with the ROSAT PSPC observation reported by Johnston et al. (1994); four sources are new.


Table 2. X-ray sources detected in the globular cluster [FORMULA]Cen ([FORMULA], [FORMULA] kpc, Djorgovski 1993) with the ROSAT HRI. Sources with a number less than 17 correspond to sources detected previously with the ROSAT PSPC (Johnston et al. 1994); those with a higher number are new sources. For each source we give the position, the statistical error in the position (in "), the distance to the cluster center in units of the core radius [FORMULA], the countrate with error, and where applicable the identification with sources detected with Einstein (Hertz & Grindlay 1983). The sources are ordered on declination. The positions given are those after correction for bore sight (see text). The positions of the center of the cluster (GC, Djorgovski & Meylan 1993), its core radius and half-mass radius (Trager et al. 1993), and positions of some optical objects discussed in the text are also listed; epochs are 1990.5 for positions by Cool et al. (1979) for USNO-A2, and 1995.5 for HD 116789.

The sources X 3 and X 4 are the Einstein sources A and D, respectively, identified with foreground M dwarfs by Cool et al. (1995a, see Table 2). Both stars can be found in the USNO-A2 catalogue (Monet et al. 1998, see Table 2). The new source X 18 can be identified with HD 116789; this is star TYC 8252 4627 1 in the Tycho Reference Catalogue, and thus its position and proper motion are well known (Hog et al. 1998). For this reason we use this star to determine the bore sight correction, i.e. the offset between the X-ray coordinates and the optical coordinates. The result is listed in Table 1.

This bore sight correction has been applied to the X-ray positions, and the resulting positions are given in Table 2. With a statistical accuracy for the X-ray position of X 18 of [FORMULA]", and taking into account small additional systematic errors (see Hasinger et al. 1998), we estimate the systematic error in the positions given in Table 2 to be less than [FORMULA]; for each individual source its statistical uncertainty should be added in quadrature to this systematic error. The positional accuracy can be improved if accurate astrometry of X 3/A and X 4/D is obtained, which will allow computation of their positions at epoch 1995.5.

In the ROSAT Deep Survey (Hasinger et al. 1998), an area with [FORMULA] radius contains 25 sources brighter than our approximate detection limit of [FORMULA]; we thus expect [FORMULA]1 serendipitous source within the core radius, i.e. the faintest source within the core radius, X 20, may well be a serendipitous background source. The brighter sources within the core radius probably are associated with [FORMULA]Cen. Outside the core radius a source is more likely to be a fore- or background source than a cluster member.

3.2. Sources in the cluster

The X-ray image of the inner area of [FORMULA]Cen is shown in Fig. 1. The half-mass radius of the cluster contains five sources. Source X 9 from Johnston et al. (1994) is clearly separated into two sources, which we denote X 9a and X 9b for the northern and southern source, respectively.

[FIGURE] Fig. 1. X-ray contours in the central area of [FORMULA]Cen as observed with the ROSAT HRI in 1995 July. The image was smoothed with a 2-d [FORMULA][FORMULA]5" Gaussian. The detected sources are indicated with their numbers in Table 2. The inner circle gives the core radius of the cluster, the outer circle the half-mass radius. The lower and left axes give pixel numbers for the ROSAT HRI detector, the upper and right axes right ascension and declination with respect to the cluster center. The conversion between pixel and celestial coordinates is accurate to within 2".

We have determined the countrates or upper limits for the six central sources in all ROSAT HRI observations of [FORMULA]Cen listed in Table 1. For the very short observations, no useful upper limits are obtained; the long-term lightcurves as determined from the other observations are shown for four of the central sources in Fig. 2. We also show the PSPC observation, dividing the PSPC counts for PSPC-X 9 equally between X 9a and X 9b. For the absorption towards [FORMULA]Cen and a black body spectrum of 0.6 keV the PSPC countrate is about 2.8 times the HRI countrate; for smaller reddening the PSPC-to-HRI count ratio varies rapidly, and therefore we do not show PSPC points of the foreground M dwarfs, whose absorption is unknown.

[FIGURE] Fig. 2. X-ray lightcurves of four sources in the center of [FORMULA]Cen and of two foreground M dwarfs (A, D). The HRI data are shown as [FORMULA] (detections, with 1-[FORMULA] errors), [FORMULA] (2-[FORMULA] upper limits). PSPC data (converted to estimated HRI countrates) are shown as [FORMULA]. There is marginal evidence for variability in sources X 9b and X 10; the foreground sources A and D are highly variable, with source A showing a large flare in 1996.

There is marginal evidence for variability in sources X 9b and X 10; and no evidence for variability of X 7 and X 9a. X 20 is detected only in July 1995 and in 1996, X 21 only in July 1995, and all upper limits in the other observations are compatible with the faint fluxes of these sources.

The July 1995 observation was obtained in two parts, separated by about 10 days. We have analysed the two parts separately, and find marginal variation of X 7 ([FORMULA] and [FORMULA] cts ksec-1 in the first and second half, respectively) and of X 9b ([FORMULA] and [FORMULA] cts ksec-1, respectively). In addition, virtually all counts of X 21 are from the first part of the observation. The number of counts of X 21 is too small for further subdivision.

To convert the observed countrates into X-ray luminosities, we assume a column and distance to [FORMULA]Cen as given in Table 2. For an 0.6 keV blackbody (see the analysis of the PSPC spectrum of X 9 in Johnston et al. 1994) 1 cts ksec-1 in the ROSAT HRI corresponds to [FORMULA] erg s-1 in the 0.5-2.5 keV band. The two sources X 9a and X 9b thus have about this luminosity, source X 7 is 40% brighter, and sources X 20 and X 21 are 40% fainter.

3.3. Sources not related to the cluster

The foreground dwarfs (Einstein X-ray sources A and D) are highly variable, as has been pointed out before (Koch-Miramond & Aurière 1987; Cool et al. 1995a). These sources also vary between the first and second half of the July 1995 observation. The extremely high flux of A in 1996 is due to a flare, which lasts almost a day (see Fig. 3).

[FIGURE] Fig. 3. Lightcurve of X 3/A in the 1996 observation. [FORMULA] indicate the numbers of counts collected in 100 s intervals; the horizontal lines indicate when ROSAT was collecting data.

HD 116789 is an A0V star. From its magnitude [FORMULA] and colour [FORMULA] we estimate [FORMULA] and a distance of about 310 pc. For an assumed bremsstrahlung spectrum of 1.4 keV the observed countrate corresponds to a luminosity in the 0.5-2.5 keV band of [FORMULA] erg s-1. It is not expected for an A0V star to emit such a flux; perhaps this star has a white dwarf companion which is responsible for the X-ray emission, as various other A0V stars detected with ROSAT; on the other hand, various apparently single A0V stars in the Bright Star Catalogue have been detected at similar and higher luminosities as HD 116789 (e.g. HD 17864, Hünsch et al. 1998). HD 116789 was detected with EXOSAT by Verbunt et al. (1986), who interpreted the detection as due to the ultraviolet leak of the CMA detector. The EXOSAT CMA countrate of this source, [FORMULA] cts s-1, converts to an X-ray luminosity at the distance of HD 116789 of about [FORMULA] erg s-1, much higher than the luminosity derived for this star from the ROSAT observations; we therefore still think that the EXOSAT countrate is due to the ultraviolet flux. The ultraviolet leak in the ROSAT HRI is far too small (Berghöfer et al. 1999) to explain the ROSAT detection.

3.4. Discussion

The positions of the sources within the half-mass radius of [FORMULA]Cen as given in Table 2 are more accurate than previously published positions, and may be used to search for optical counterparts. We have done this among the variables (contact binaries, detached binaries, and suspected RS CVn stars) found in [FORMULA]Cen by Kaluzny et al. (1996, 1997): no counterpart is among these stars. (Only one of these variables is in the area shown in Fig. 1, viz. the contact binary OGLEGC 13.) Our non-detection of these binaries is not surprising, considering that our detection limit is above [FORMULA] erg s-1: all of the contact binaries hitherto detected in X-rays (McGale et al. 1996), and many RS CVn systems (Dempsey et al. 1993) are less luminous than this.

Cool et al. (1995a) argue that X 7/B is an extended source. This source is detected in the ROSAT PSPC observation (Johnston et al. 1994) and in the ROSAT HRI observations of 1994, 1995 January and July, and 1996. All of these observations are more sensitive than the 1992 and 1993 observations used by Cool et al. (1995a); in all of them X 7/B is compatible with being a point source.

With the identification of sources X 3 and X 4 with foreground stars, we can reinvestigate the suggested identification of X 5/E with a foreground K star, as suggested by Margon & Bolte (1987). We use the optical positions of A and D to determine the offset between the X-ray positions of the PSPC observation as listed in Johnston et al. (1994) and the optical positions. We then apply this offset to the position of X 5, and find that the resulting position is at 6[FORMULA]5 arcseconds from the optical star. We take the position of the optical star (Table 2) from USNO A2 0375-18334783 (Monet et al. 1998). Identification of X 5 with the foreground star is therefore a distinct possibility.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: June 20, 2000