2 X-ray sources inside the Lagoon Nebula

The Lagoon Nebula was observed serendipitously for about 20 ksec with the XMM-Newton observatory (Jansen et al. 2001) during our AO1 observation of the O4 V star 9 Sgr (JD 2451976.978 - 2451977.239). The two EPIC-MOS instruments were operated in the full frame mode (Turner et al. 2001) whilst the EPIC-pn camera was used in the extended full frame mode (Strüder et al. 2001). All three EPIC instruments used the thick filter to reject optical light. We used version 5.1 of the XMM-Newton Science Analysis System ( SAS) to reduce the raw EPIC data. More details on the pipeline processing of the data are given in the paper dealing with the 9 Sgr data (Rauw et al. 2002).

The combined EPIC soft band (0.5-1.2 keV) X-ray image around 9 Sgr reveals a number of point-like (and maybe also diffuse) sources (see Fig. 1). Figure 2 displays the EPIC contours of the X-ray sources superimposed on a Digital Sky Survey image of M 8. This image indicates that most of the fainter sources are located in the region of the very young open cluster NGC 6530. One rather bright and apparently diffuse source (to the south-west of 9 Sgr) is associated with the Hourglass Region.

Because of the contamination of the EPIC data by hard straylight photons from the nearby LMXB GX 9 +1, the source detection was performed on the soft band images (i.e. where there is no contamination) only. For details on the properties of the straylight, we refer again to the work of Rauw et al. (2002).

Using the SAS source detection algorithms (see e.g. Hasinger et al. 2001), we find 250 sources that are detected simultaneously in the MOS1, MOS2 and pn images with a combined logarithmic likelihood

$$\sum_{i = 1}^{3} -\ln{p_i} \geq 9.0$$

where the summation runs over the MOS1, MOS2 and pn images. In this formula, p_i stands for the probability that a random Poissonian fluctuation may have caused the observed source counts in the image of the detector i. We then inspected each source manually to reject spurious detections due e.g. to hot pixels. This resulted in a total of 220 confirmed detections. Out of these 220 sources, 117 have a combined likelihood $\geq$20.0 and are detected in the individual images from all three EPIC instruments with $-\ln{p_i} \geq 3.0$. With these criteria, the probability of detecting a spurious source within a $15\hbox{$^{\prime\prime}$ }\times 15\hbox{$^{\prime\prime}$ }$ detection element should be $\sim$ $2 \times 10^{-9}$ and we therefore expect much less than one spurious detection over the entire EPIC field of view. To this list we add two sources that are not detected in at least one of the instruments because they fall outside its field of view, but are clearly detected in the remaining instruments. This gives us a total of 119 highly significant detections. In the following, we will restrict our discussion to these latter objects (see Fig. 3). The faintest sources in this category have about 10^-3 cts s^-1 over the 0.5-1.2 keV band of the MOS instruments. Assuming a 1 keV thermal spectrum with a neutral hydrogen column density of $0.17 \times 10^{22}$ cm^-2 (see also Sect. 3 below), the faintest sources correspond to an observed flux of about $8.9 \times 10^{-15}$ erg cm^-2 s^-1 and an unabsorbed flux of $13.1 \times 10^{-15}$ erg cm^-2 s^-1 in the 0.5-5.0 keV energy range. Note that the corresponding observed flux in the 0.5-1.2 keV soft band would be $5.2 \times 10^{-15}$ erg cm^-2 s^-1.

Figure 2: EPIC X-ray contours (corresponding to the combined MOS1 + MOS2 + pn data in the 0.5-1.2 keV range) superimposed on a DSS optical image of the Lagoon Nebula. The different contours correspond to 5, 15, 24, 50 and 200 counts in the X-ray image.

Figure 3: DSS optical image of the Lagoon Nebula complex together with the positions of the highly significant sources. Open circles and crosses stand for X-ray sources having respectively a single or several counterparts in the Sung et al. (2000) catalogue. The open triangles indicate the positions of sources with no counterpart in the latter catalogue (nor in the SIMBAD database).

The most complete compilation of optical sources in NGC 6530 is the catalogue provided by Sung et al. (2000). These authors present UBVRI and H$\alpha$ photometry of 887 stars brighter than V = 17 and spread over an area from RA = 18:03:20 to 18:04:52 and from DEC = -24:11:45 to -24:32:20 (Equinox J2000.0). They include also a deeper HST WFPC2 observation of the Hourglass Region in their analysis.

We find that a total of 111 out of 119 X-ray sources detected with XMM fall inside the area covered by Sung et al. (2000). We have cross-correlated the positions of these sources with the Sung et al. catalogue. In order to determine the optimal radius of cross-correlation, we adopted the approach outlined by Jeffries et al. (1997). We generated the cumulative distribution of the number of detected sources as a function of the cross-correlation radius r (see Fig. 4) and modelled this distribution assuming that it can be represented by means of an expression taken from Jeffries et al.:

$$\Phi(d \leq r) = A~\left[1 - \exp{\left(\frac{-r^2}{2~\sigma^2}\right)}\right] + (N - A)~\left[1 - \exp(-\pi~B~r^2)\right].$$

Here N, A, $\sigma$ and B stand for the total number of cross-correlated X-ray sources (N = 111), the number of true correlations, the uncertainty on the X-ray source position and the surface density of optical sources respectively. The first term in this expression stands for the cumulative distribution of true correlations whereas the second term yields the cumulative number of spurious correlations. A, $\sigma$ and B are fitting parameters and were obtained from the best fit to the distribution displayed in Fig. 4. We obtain A = 87.0, $\sigma = 2.9$ arcsec and $B = 4.7 \times 10^{-4}$ arcsec^-2. The optimal correlation radius, i.e. the radius that includes the bulk of the true correlations while simultaneously limiting contamination by spurious correlations, is found to be around 9 arcsec. For r = 9 arcsec, we expect to achieve 86 true and only 3 spurious correlations.

There are two limitations to this procedure. Strictly speaking, the above expression for the spurious detections is downright valid for a uniform distribution of the optical sources over the investigated area. Though this is not the case for the optically brightest objects, Fig. 1 of Sung et al. (2000) indicates that this should be a fairly good approximation for the fainter stars. Another issue is that we assume that a single correlation radius can be used over the entire field of view. A radius of 9 arcsec corresponds roughly to 1.5 times the in-orbit FWHM of the on-axis point spread function (PSF) of the XMM mirror modules at 1.5 keV (Jansen et al. 2001). At large off-axis angles, the quality of the PSF progressively degrades and the half energy width increases by nearly a factor of two (Stockman et al. 1998). Our optimal radius should therefore be seen as some sort of weighted average for different off-axis angles.

Within a correlation radius of 9 arcsec, we find that:

$\bullet$

the brightest X-ray source is associated with the O4 V((f⁺)) star 9 Sgr. The data on this star have been analysed elsewhere (Rauw et al. 2002) and we will not repeat this discussion here;

$\bullet$

a probably diffuse X-ray source is associated with the Hourglass Region. We shall return to the HG in Sect. 4;

$\bullet$

72 X-ray sources have a single optical counterpart within a radius of less than 9 arcsec. The average angular separation between the X-ray source and the optical counterpart is $(4.1 \pm 2.1)$ arcsec. The brightest soft X-ray sources in this category are HD 164816 (O9.5 III-IV) and SCB 731. The latter is a very red 16th magnitude star without outstanding features (such as H$\alpha$ emission) in its optical photometry. The X-ray source associated with SCB 731 appears elongated since it lies rather far off-axis;

$\bullet$

17 X-ray sources have two or more optical stars falling within the 9 arcsec radius. Seven of these sources have at least one known H$\alpha$ emission star inside their "error box'';

$\bullet$

28 sources do not have an optical counterpart in the catalogue of Sung et al. We have cross-correlated these sources with the Guide Star Catalogue (GSC). In most cases, we find one or several GSC objects inside the 9 arcsec radius. These optical counterparts are usually very faint (V or $R \geq 17$) except for sources # 91, 92 and 94 that have counterparts with R magnitudes 15.0, 14.6 and 13.3 respectively (note that these sources fall outside the area investigated by Sung et al.). Sources # 93, 96, 98, 99, 100, 101, 102 and 112 have no GSC counterpart. Most of the objects in Table 3 are therefore X-ray sources with high L$_{\rm X}$/L $_{\rm vis}$ ratios. Given the galactic coordinates of NGC 6530 ( $l_{\rm II} = 6.14^{\circ}$, $b_{\rm II} = -1.38^{\circ}$), the total galactic column density along our line of sight in this direction must be extremely large and the number of extragalactic sources in our soft detection energy band should be extremely low.

 

 

Table 1: X-ray sources detected in NGC 6530 and found to have a single counterpart in the catalogue of Sung et al. (2000) or in the SIMBAD database. The first and second columns yield the number of the X-ray source as well as the name according to the conventions for serendipitous XMM-Newton sources. Columns [3]-[5] provide the vignetting and background corrected EPIC-MOS and pn count rates in the 0.5-1.2 keV band. For sources affected by the gaps between the detectors, we do not quote the count rate for the corresponding instrument. The number of the optical counterpart from Sung et al. (2000) and the separation from the position of the X-ray source are given in Cols. [6] and [7]. The photometric data are taken from Sung et al. (2000), except for V 1752 Sgr. Column [11] indicates whether variability is detected at the 95% confidence level (Y) or not (N). Those sources flagged with a "?'' show some hints of variability in their light curve, but this is not found to be significant by the Kolmogorov-Smirnov test. Finally, the last column yields information about the spectral types, pre-main sequence classification and the membership probability ($P_{\mu }$) according to the proper motion study of van Altena & Jones (1972).

Source	XMMU J	MOS1	MOS2	pn	SCB	d	V	B-V	V-I	Variability	Comments
#		10-3 cts s-1	10-3 cts s-1	10-3 cts s-1		(arcsec)
[1]	[2]	[3]	[4]	[5]	[6]	[7]	[8]	[9]	[10]	[11]	[12]
1	180253.7-242053	4.1 $\pm$ 1.0	2.8 $\pm$ 1.1	5.2 $\pm$ 1.6		8.4	15.80				V 1752 Sgr = PMS
2	180322.8-242208	3.0 $\pm$ 0.6	1.4 $\pm$ 0.4		7	3.9	10.62	0.96	1.07	N
3	180323.3-241525	1.4 $\pm$ 0.7	1.5 $\pm$ 0.7	5.2 $\pm$ 1.0	20	7.0	15.17	1.28	1.53
4	180323.8-241556	2.7 $\pm$ 0.6	3.1 $\pm$ 0.6	2.0 $\pm$ 0.8	28	5.7	15.51	1.25	1.52
5	180326.9-241246	2.9 $\pm$ 0.7	2.1 $\pm$ 0.6	6.2 $\pm$ 1.2	47	1.4	11.83	0.26	0.45		$P_{\mu} = 0.80$
6	180339.8-242200	1.6 $\pm$ 0.4	1.7 $\pm$ 0.4	2.8 $\pm$ 0.7	1032	8.8	18.72		2.98		PMS
7	180340.1-241923	1.6 $\pm$ 0.4	1.7 $\pm$ 0.4	2.8 $\pm$ 0.7	138	3.5	15.05	1.24	1.41
8	180341.5-241222	1.5 $\pm$ 0.5	1.0 $\pm$ 0.4	1.7 $\pm$ 0.7	151	0.8	15.38	1.33	1.58
9	180341.5-242416	1.9 $\pm$ 0.4	2.2 $\pm$ 0.4	4.3 $\pm$ 0.7	157	5.7	15.44	1.47	1.93		PMS
10	180343.4-242340	2.6 $\pm$ 0.5	2.0 $\pm$ 0.4	3.8 $\pm$ 1.0	174	4.1	16.90	1.44	1.88
11	180345.1-242205	5.1 $\pm$ 0.6	7.3 $\pm$ 0.7	11.5 $\pm$ 1.2	182	0.5	9.95	0.18	0.41	?	B2.5V, $P_{\mu} = 0.86$
12	180347.0-241647	0.8 $\pm$ 0.3	1.6 $\pm$ 0.4	1.6 $\pm$ 0.5	199	1.8	10.75	0.09	0.15
13	180347.2-242201	1.0 $\pm$ 0.6	1.8 $\pm$ 0.7	3.8 $\pm$ 0.9	202	4.5	16.64	1.55	2.13
14	180349.1-241537	0.7 $\pm$ 0.3	0.8 $\pm$ 0.3	1.7 $\pm$ 0.6	220	4.1	16.87	1.31	1.55
15	180349.2-242343	0.8 $\pm$ 0.4	0.7 $\pm$ 0.3	2.4 $\pm$ 0.6	223	6.0	16.85	1.35	1.72
16	180350.9-242018	2.5 $\pm$ 0.5	1.7 $\pm$ 0.4	3.6 $\pm$ 0.9	238	5.6	16.96	1.44	2.07
17	180352.6-242256	4.3 $\pm$ 0.6	4.6 $\pm$ 0.6	8.7 $\pm$ 1.2	253	6.4	13.80	1.23	1.49	N	$P_{\mu} = 0.00$
18	180355.4-241701	2.8 $\pm$ 0.5	4.7 $\pm$ 0.6	7.0 $\pm$ 1.1	274	1.8	14.61	1.06	1.18	N
19	180356.6-241845	14.3 $\pm$ 1.0	16.6 $\pm$ 1.1	35.8 $\pm$ 2.0	282	4.1	6.99	0.02	0.00	N	HD 164816 (O9.5 III-IV)
20	180357.3-241611	1.3 $\pm$ 0.4	1.3 $\pm$ 0.4	4.2 $\pm$ 0.8	285	1.2	16.27	1.20	1.41
21	180358.3-242455	3.7 $\pm$ 0.5	3.4 $\pm$ 0.5	3.7 $\pm$ 0.9	295	5.6	15.51	1.40	1.84	N
22	180400.2-241915	3.2 $\pm$ 0.5	4.0 $\pm$ 0.6	6.9 $\pm$ 1.0	314	3.4	15.28	1.14	1.36	N
23	180402.9-242140	4.4 $\pm$ 0.6	5.4 $\pm$ 0.6	5.3 $\pm$ 0.9	340	0.1	14.85	1.35	1.59	N
24	180404.9-241643	1.3 $\pm$ 0.4	0.7 $\pm$ 0.3	2.5 $\pm$ 0.6	360	4.3	16.31	1.40	1.64
25	180407.2-241926	1.5 $\pm$ 0.5	1.3 $\pm$ 0.4	3.6 $\pm$ 0.7	379	2.7	10.76	0.25	0.48		$P_{\mu} = 0.00$
26	180407.9-242126	1.7 $\pm$ 0.4	2.0 $\pm$ 0.4	3.5 $\pm$ 0.8	387	0.6	16.43	1.32	1.68
27	180409.4-242715	2.4 $\pm$ 0.5	2.1 $\pm$ 0.5	3.8 $\pm$ 0.8	410	6.2	16.69	1.60	1.86
28	180411.0-242149	2.0 $\pm$ 0.4	1.6 $\pm$ 0.4	3.8 $\pm$ 0.8	426	4.7	10.37	0.12	0.23		B3Ve, $P_{\mu} = 0.84$
29	180411.8-242527	1.7 $\pm$ 0.4	2.6 $\pm$ 0.5	4.2 $\pm$ 0.9	432	4.1	11.69	0.24	0.44		A0, $P_{\mu} = 0.52$
30	180412.5-241950	4.0 $\pm$ 0.6	2.9 $\pm$ 0.5	11.1 $\pm$ 1.5	440	7.0	15.99	1.32	1.70	N	PMS
31	180413.0-241825	1.8 $\pm$ 0.4	1.7 $\pm$ 0.4	3.8 $\pm$ 0.9	450	2.7	14.34	0.98	1.22
32	180413.2-242452	1.8 $\pm$ 0.4	1.8 $\pm$ 0.4	4.0 $\pm$ 0.9	452	2.8	15.31	1.35	1.65
33	180413.8-242619	1.4 $\pm$ 0.5	1.3 $\pm$ 0.5	5.3 $\pm$ 0.9	457	5.4	16.32	1.53	1.98
34	180413.9-242405	2.6 $\pm$ 0.5	2.6 $\pm$ 0.5	3.8 $\pm$ 1.0	454	1.4	15.59	1.40	1.68

 

Table 1: continued.

Source	XMMU J	MOS1	MOS2	pn	SCB	d	V	B-V	V-I	Variability	Comments
#		10-3 cts s-1	10-3 cts s-1	10-3 cts s-1		(arcsec)
[1]	[2]	[3]	[4]	[5]	[6]	[7]	[8]	[9]	[10]	[11]	[12]
35	180414.3-241938	0.8 $\pm$ 0.5	1.3 $\pm$ 0.5	4.3 $\pm$ 1.0	465	3.0	14.39	1.22	1.46
36	180414.4-242155	7.6 $\pm$ 0.7	7.2 $\pm$ 0.8	15.9 $\pm$ 1.7	470	4.2	13.45	1.20	1.48	N	$P_{\mu} = 0.00$
37	180414.4-242332	2.2 $\pm$ 0.6	1.8 $\pm$ 0.6	7.8 $\pm$ 1.2	458	7.9	14.61	1.08	1.29
38	180415.1-242112	1.6 $\pm$ 0.4	2.5 $\pm$ 0.5	3.3 $\pm$ 1.1	480	8.3	16.83	1.49	1.90
39	180415.8-241457	1.9 $\pm$ 0.5	1.3 $\pm$ 0.5	3.9 $\pm$ 1.0	481	1.8	12.79	0.27	0.42		$P_{\mu} = 0.76$
40	180416.8-242835	2.4 $\pm$ 0.6	1.7 $\pm$ 0.5	4.4 $\pm$ 1.0	502	2.3	11.56	0.23	0.42		A0, $P_{\mu} = 0.60$
41	180417.5-242415	2.0 $\pm$ 0.5	3.2 $\pm$ 0.5	2.4 $\pm$ 0.9	504	6.8	14.26	1.22	1.44
42	180420.2-242250	2.7 $\pm$ 0.5	2.8 $\pm$ 0.5	6.0 $\pm$ 1.1	540	2.3	16.17	1.26	1.76		PMS
43	180420.4-242204	1.2 $\pm$ 0.5	1.6 $\pm$ 0.6	6.8 $\pm$ 1.1	546	3.4	15.43	1.33	1.58
44	180420.9-242324	2.7 $\pm$ 0.5	4.3 $\pm$ 0.6	10.8 $\pm$ 1.3	556	4.3	16.98	1.41	2.11	N	PMS
45	180421.1-241951	4.2 $\pm$ 0.6	4.1 $\pm$ 0.6	11.2 $\pm$ 1.4	557	2.7	14.71	1.11	1.44	N
46	180421.8-242039	4.2 $\pm$ 0.6	4.7 $\pm$ 0.7	13.4 $\pm$ 1.6	571	5.9	13.92	0.84	1.06	N
47	180422.6-242355	5.2 $\pm$ 0.7	6.2 $\pm$ 0.8		586	4.4	16.81	1.47	1.82	N
48	180423.6-242951	1.4 $\pm$ 0.4	0.8 $\pm$ 0.5	2.3 $\pm$ 0.8	592	4.3	16.76	1.42	1.79
49	180424.2-241829	3.2 $\pm$ 0.6	1.7 $\pm$ 0.5	4.7 $\pm$ 0.9	600	2.6	14.19	1.04	1.34
50	180424.5-242025	2.6 $\pm$ 0.5	2.4 $\pm$ 0.5	6.7 $\pm$ 1.2	607	3.1	16.47	1.45	1.80
51	180425.0-242747	2.6 $\pm$ 0.6	1.3 $\pm$ 0.6	7.7 $\pm$ 1.3	611	2.0	11.43	0.21	0.41		$P_{\mu} = 0.70$
52	180425.7-242254	1.9 $\pm$ 0.7	3.7 $\pm$ 0.9	8.8 $\pm$ 1.5	614	6.1	16.32		1.88
53	180425.8-241750	1.3 $\pm$ 0.5	1.3 $\pm$ 0.5	2.4 $\pm$ 0.9	617	2.7	15.36	1.22	1.50
54	180425.8-242357	2.8 $\pm$ 0.8	2.3 $\pm$ 0.7	6.3 $\pm$ 1.3	627	8.4	16.69	1.43	1.93
55	180427.5-242354	1.9 $\pm$ 0.6	2.0 $\pm$ 0.6	3.6 $\pm$ 1.1	641	2.5	13.77	1.09	1.33
56	180428.6-241855	1.9 $\pm$ 0.6	3.9 $\pm$ 0.7	6.9 $\pm$ 1.3	654	2.3	14.17	0.97	1.19	N
57	180428.7-241400	1.1 $\pm$ 0.6	2.6 $\pm$ 0.8	6.6 $\pm$ 1.5	656	3.8	11.99	0.52	0.70		$P_{\mu} = 0.83$
58	180428.7-241444	1.7 $\pm$ 0.5	1.2 $\pm$ 0.5	6.9 $\pm$ 1.4	655	3.0	15.39	0.63	1.52
59	180429.0-242225	5.1 $\pm$ 0.7	6.5 $\pm$ 0.8		658	6.5	15.12	1.26	1.57	N
60	180429.5-242021	3.7 $\pm$ 0.6	2.8 $\pm$ 0.6	7.3 $\pm$ 1.4	665	6.8	16.86	1.54	2.10	N
61	180431.9-242406	2.6 $\pm$ 0.6	2.7 $\pm$ 0.6	6.0 $\pm$ 1.3	700	7.1	14.59	1.18	1.46
62	180432.4-242122	2.9 $\pm$ 0.6	2.5 $\pm$ 0.6	3.1 $\pm$ 1.3	698	6.5	16.11	1.41	1.79
63	180432.9-241844	3.4 $\pm$ 0.7	3.1 $\pm$ 0.6	9.8 $\pm$ 1.6	707	1.5	10.42	0.12	0.19	N	B2.5Vne, $P_{\mu} = 0.75$
64	180433.6-242706	6.8 $\pm$ 1.0	3.5 $\pm$ 0.7	7.6 $\pm$ 2.0	718	4.2	14.56	1.23	1.57	N
65	180434.3-241925	2.2 $\pm$ 0.6	3.0 $\pm$ 0.6	6.3 $\pm$ 1.3	725	3.1	12.85	0.87	1.02		$P_{\mu} = 0.01$
66	180435.5-242729	15.5 $\pm$ 1.5	15.7 $\pm$ 1.5	50.9 $\pm$ 3.7	731	2.0	16.53	1.50	2.03	Y
67	180437.1-242956	2.2 $\pm$ 0.6	1.2 $\pm$ 0.6	5.2 $\pm$ 1.4	747	2.7	15.02	1.40	1.63
68	180438.1-242043	1.9 $\pm$ 0.6	2.1 $\pm$ 0.6	4.0 $\pm$ 1.2	761	4.1	15.41	1.40	1.74
69	180439.1-242409	1.4 $\pm$ 0.7	2.3 $\pm$ 0.7	6.3 $\pm$ 1.4	763	5.7	10.21	1.21	1.27		K0III, $P_{\mu} = 0.00$
70	180441.8-241855	2.5 $\pm$ 0.7	2.6 $\pm$ 0.7	6.4 $\pm$ 1.4	786	4.1	14.62	1.04	1.24
71	180441.8-242050	1.4 $\pm$ 0.5		2.8 $\pm$ 0.9	787	5.9	9.76	0.09	0.18		HD 164947, B2.5 V
72	180450.6-242540	4.3 $\pm$ 1.0			879	2.5	11.94	0.39	0.75		LkH$\alpha$ 115 (B4Ve), $P_{\mu} = 0.79$

   

Table 2: Same as Table 1 for X-ray sources with two or more counterparts from Sung et al. (2000) inside a radius of 9 arcsec around the position of the X-ray source.

Source	XMMU J	MOS1	MOS2	pn	SCB	d	Variability	Comments
#		10-3 cts s-1	10-3 cts s-1	10-3 cts s-1		(arcsec)
[1]	[2]	[3]	[4]	[5]	[6]	[7]	[8]	[9]
73	180338.1-242235	1.8 $\pm$ 0.5	1.3 $\pm$ 0.5	4.4 $\pm$ 0.9	1039/1028	1.7/3.3		a
74	180340.3-242316	4.9 $\pm$ 0.6	5.1 $\pm$ 0.8	9.7 $\pm$ 1.3	142/146/1019	4.2/5.5/6.8		b
75	180350.5-242110				240/228/229			240 = LkH$\alpha$108c
76	180407.4-242226	7.8 $\pm$ 0.8	7.8 $\pm$ 0.8		382/380	3.7/4.0	N
77	180407.7-242311	1.7 $\pm$ 0.4	1.4 $\pm$ 0.4	1.8 $\pm$ 0.7	388/390	2.8/3.2		390 = PMS
78	180414.7-242224	1.5 $\pm$ 0.6	3.2 $\pm$ 0.6		468/1005	3.2/3.3		1005 = PMS
79	180415.3-241902	0.8 $\pm$ 0.4	1.7 $\pm$ 0.5	2.6 $\pm$ 0.9	478/482	5.0/6.9		482 = PMSc
80	180421.8-242114	1.4 $\pm$ 0.8	1.8 $\pm$ 0.8	9.7 $\pm$ 1.4	573/561/558	5.5/8.0/8.4
81	180423.0-242116	7.5 $\pm$ 0.8	8.3 $\pm$ 0.9	14.5 $\pm$ 2.1	582/593/587	6.5/8.3/8.7	N
82	180426.9-242100	2.0 $\pm$ 0.5	2.3 $\pm$ 0.5	4.5 $\pm$ 1.1	631/635	7.1/7.2
83	180429.1-242147	1.5 $\pm$ 0.6	2.0 $\pm$ 0.7	4.5 $\pm$ 1.5	657/663	5.3/6.1
84	180429.5-242300	7.7 $\pm$ 0.9	7.0 $\pm$ 0.9	25.4 $\pm$ 2.3	671/659	4.4/7.2	N
85	180430.7-242634	4.6 $\pm$ 0.8	3.7 $\pm$ 0.7	7.8 $\pm$ 1.4	681/682	2.9/4.1	N
86	180434.2-242154	1.7 $\pm$ 0.5	1.6 $\pm$ 0.6	2.8 $\pm$ 1.0	723/720/716	4.1/7.0/8.3
87	180436.0-241959	1.0 $\pm$ 0.6	2.3 $\pm$ 0.7	3.4 $\pm$ 1.0	740/738	3.5/6.8
88	180439.0-241924	2.0 $\pm$ 0.6	2.0 $\pm$ 0.6	3.4 $\pm$ 1.0	765/773	1.9/8.6
89	180448.7-242634	10.5 $\pm$ 1.4	9.3 $\pm$ 1.4		864/862	1.9/6.8	N	862 = PMS

^a SCB 1028 is classified as a PMS star, while SCB 1039 is a PMS candidate (Sung et al. 2000).
^b SCB 142 is a H$\alpha$ emitting star (LkH$\alpha$106) while SCB 1019 is a PMS star (Sung et al.).
^c Source #75 (XMMU J180350.5-242110) lies within the wings of the PSF of 9 Sgr. Therefore no accurate count rates can be provided.

The sources of the last three categories are listed in Tables 1-3 respectively. Along with our own source numbering, we list the source designation following the naming conventions recommended by the XMM SOC and the IAU (see the XMM-Newton Newsletter #4, April 2001): the XMMU J prefix is followed by the right ascension HHMMSS.s (in hours, minutes, seconds and tenths of seconds) and the declination of the source $\pm$DDMMSS (in degrees, arcminutes and arcseconds). Note that the coordinates refer to equinox J2000 and are truncated, not rounded.

Figure 4: Cumulative numbers of correlations between the X-ray detections and the Sung et al. (2000) catalogue as a function of correlation radius. The dotted curves correspond to the best fitting expressions for the real and spurious correlations. The dashed curve yields the sum of these terms and the dash-dotted vertical line corresponds to the optimal correlation radius of 9 arcsec.

2.1 Light curves

We have extracted light curves and spectra for the brightest objects. Except for sources near the edge of the field of view, we used a circular extraction region with a radius between 19 and 40 arcsec depending on the angular separation between neighbouring X-ray sources. For sources near the edge of the field of view, we used elliptical extraction regions to match as closely as possible the shape of the point spread function. The light curves and spectra were accumulated over the PI range 200 to 10000 (corresponding roughly to $E \in [0.2, 10.0]$ keV), except for objects that are contaminated by the straylight. In the latter cases, we extracted light curves only (i.e. no spectra) over the PI range 200 to 1500 (i.e. $E \in [0.2, 1.5]$ keV).

For all the light curves we adopted time bins of 400 s. The light curves were tested for variability using a Kolmogorov-Smirnov test against a model of constant count rate. Among the sources with a single optical counterpart, SCB 731, the second brightest X-ray source, is clearly variable: the light curve (extracted over an elliptical region adapted to the shape of the PSF) reveals a strong X-ray flare towards the end of our observation (Fig. 5). The X-ray flux suddenly rises by nearly two orders of magnitude and after about 2000 s starts to decline again.

SCB 182 and source # 99 show some marginal variability in their light curves. The behaviour is the same in all three EPIC instruments, but the variability occurs at a rather low level so that it is found not to be significant by the Kolmogorov-Smirnov test (at the $P > 95\%$ confidence level). A Fourier technique detects slow variability in the light curve of source # 99 at the 99% confidence level. None of the other relatively bright sources turned out to display significant variability (95% level) over the duration of our observation.

   

Table 3: X-ray sources detected in NGC 6530 that have no counterpart in the Sung et al. (2000) catalogue and the SIMBAD database within a radius of 9 arcsec.

Source	XMMU J	MOS1	MOS2	pn	Variability
#		10-3 cts s-1	10-3 cts s-1	10-3 cts s-1
[1]	[2]	[3]	[4]	[5]	[6]
90	180252.9-241525	a	a	11.4 $\pm$ 2.3
91	180253.3-241955	6.4 $\pm$ 1.4		12.7 $\pm$ 2.3	N
92	180257.3-242014	3.9 $\pm$ 1.1	2.9 $\pm$ 1.1	10.4 $\pm$ 1.7
93	180337.5-242116	0.5 $\pm$ 0.4	1.1 $\pm$ 0.4	2.4 $\pm$ 0.7
94	180339.1-243302	5.1 $\pm$ 1.0	6.9 $\pm$ 1.1	12.9 $\pm$ 1.9	N
95	180351.3-241149	1.8 $\pm$ 0.5	1.8 $\pm$ 0.6	1.6 $\pm$ 0.8
96	180356.4-242340	1.3 $\pm$ 0.4	2.3 $\pm$ 0.4	3.9 $\pm$ 0.8
97	180356.5-242018	0.8 $\pm$ 0.4	0.4 $\pm$ 0.4	1.6 $\pm$ 0.8
98	180357.1-242055	1.8 $\pm$ 0.4	1.4 $\pm$ 0.5	1.8 $\pm$ 0.7
99	180358.6-242530	6.4 $\pm$ 0.7	6.5 $\pm$ 0.7	16.0 $\pm$ 1.6	?
100	180402.6-242645	2.3 $\pm$ 0.5	2.2 $\pm$ 0.4
101	180404.5-241941	0.7 $\pm$ 0.3	1.2 $\pm$ 0.4	1.2 $\pm$ 0.5
102	180405.0-242511	2.9 $\pm$ 0.5	3.4 $\pm$ 0.5	9.1 $\pm$ 1.1
103	180405.2-242340	0.8 $\pm$ 0.3	1.1 $\pm$ 0.4	1.1 $\pm$ 0.5
104	180409.3-241622	1.2 $\pm$ 0.4	0.8 $\pm$ 0.4	2.6 $\pm$ 0.8
105	180409.3-242017	1.7 $\pm$ 0.5	1.4 $\pm$ 0.4	2.6 $\pm$ 0.8
106	180412.2-241907	1.5 $\pm$ 0.5	1.3 $\pm$ 0.5	1.9 $\pm$ 0.8
107	180413.2-241517	1.4 $\pm$ 0.5	1.0 $\pm$ 0.5	2.2 $\pm$ 0.8
108	180413.9-241722	0.8 $\pm$ 0.4	1.3 $\pm$ 0.5	3.2 $\pm$ 0.9
109	180419.6-242603	1.1 $\pm$ 0.4	1.1 $\pm$ 0.4	3.6 $\pm$ 1.0
110	180420.3-241920	0.9 $\pm$ 0.4	1.6 $\pm$ 0.5	2.6 $\pm$ 1.0
111	180420.4-242810	2.3 $\pm$ 0.6	2.4 $\pm$ 0.6	5.0 $\pm$ 1.2
112	180424.7-242454	1.9 $\pm$ 0.6	1.5 $\pm$ 0.5	3.0 $\pm$ 0.8
113	180430.5-242850	1.2 $\pm$ 0.5	1.9 $\pm$ 0.6	3.4 $\pm$ 1.1
114	180435.0-242334	3.3 $\pm$ 0.7	4.2 $\pm$ 0.7	10.0 $\pm$ 1.6	N
115	180440.3-242643	7.3 $\pm$ 1.1	6.9 $\pm$ 1.1	19.0 $\pm$ 2.4	N
116	180441.9-242248	2.0 $\pm$ 0.7	1.9 $\pm$ 0.7	7.8 $\pm$ 1.4
117	180441.9-242128	3.8 $\pm$ 0.9	1.7 $\pm$ 0.7	12.5 $\pm$ 1.8

^a Source #90 (XMMU J180252.9-241525) lies outside the field of view of the MOS instruments, but is detected as a rather bright source with $(1.14 \pm 0.23) \times 10^{-2}$ cts s^-1 in the pn image.

Figure 5: EPIC light curves of the X-ray source # 66 (XMMU J180435.5-242729) associated with SCB 731. We have also extracted the light curve of a nearby background region (not shown here). The latter does not display any variability. The zero time in our light curves corresponds to JD 2451976.982.

We also reduced a ROSAT-PSPC observation of M 8 (rp900374n00, integration time 10.4 ksec, JD 2449078.532 - 2449079.569) using the XSELECT software. In this exposure, SCB 253 appears as the third brightest source (after 9 Sgr and HD 164816) with a net count rate of $(6.9 \pm 1.0) \times 10^{-3}$ cts s^-1 over the entire energy range of the PSPC instrument. Other sources that are detected in this PSPC image are the Hourglass Region, SCB 7, 28, 47, 182, 274, 340, 452/454, 470, #76, #95, #91, #94 as well as a source which is not seen with XMM at RA = 18:04:03.2, DEC = -24:31:08.5 (equinox J2000).

2.2 Spectra

For the spectra, we adopted the redistribution matrices provided by the EPIC instrument teams (versions available in May 2001) and we used the SAS to build the appropriate ancillary response files for each EPIC instrument. The spectra were binned to reach a minimum of 10 counts per channel and the background corrected spectra were analyzed using the XSPEC software (version 11.00). Because of the strong noise in the pn detector below 0.2 keV and the uncertain calibration of the EPIC instruments at low energies, we ignored the binned energy channels below 0.3 keV for the XSPEC fits. Note that for the X-ray sources associated with HD 164816 and SCB 731, we have a total of respectively 300 and 800 net counts for each MOS spectrum. The pn spectra contain about twice as many counts.

 

 

Table 4: Best fitting parameters for absorbed thermal plasma mekal and absorbed power law models fitted to the EPIC spectra of the brightest X-ray sources in NGC 6530. The observed fluxes quoted in Cols. [5] and [9] are evaluated over the energy range 0.5-5 keV. The quoted uncertainties correspond to formal errors on the spectral fits.

	wabs*mekal					wabs*power
Source #	$N_{\rm H}$	kT	$\chi^2_{\nu}$ (d.o.f.)	flux		$N_{\rm H}$	$\Gamma$	$\chi^2_{\nu}$ (d.o.f.)	flux
	1022 cm-2	keV		erg cm-2 s-1		1022 cm-2			erg cm-2 s-1
[1]	[2]	[3]	[4]	[5]		[6]	[7]	[8]	[9]
11	0.22+.07-.05	3.18+.66-.55	0.98 (60)	$1.09 \times 10^{-13}$		0.43+.12-.13	2.49+.30-.26	1.05 (60)	$1.07 \times 10^{-13}$
19	0.43+.12-.13	0.23+.04-.04	1.08 (152)	$0.77 \times 10^{-13}$
36	0.16+.08-.06	3.05+.65-.53	1.30 (59)	$1.09 \times 10^{-13}$		0.49+.11-.09	2.78+.33-.29	1.07 (59)	$1.08 \times 10^{-13}$
47	0.15+.06-.04	2.95+.66-.41	1.02 (61)	$1.17 \times 10^{-13}$		0.43+.12-.10	2.92+.46-.37	0.82 (61)	$1.13 \times 10^{-13}$
59	1.06+.16-.19	0.70+.27-.10	1.55 (40)	$0.65 \times 10^{-13}$		0.42+.14-.11	2.99+.54-.40	1.01 (40)	$0.83 \times 10^{-13}$
66	0.26+.04-.04	6.64+2.27-1.35	0.82 (130)	$5.24 \times 10^{-13}$		0.34+.07-.06	1.80+.15-.15	0.78 (130)	$5.27 \times 10^{-13}$
76	0.11+.04-.04	3.54+.91-.60	1.67 (65)	$1.05 \times 10^{-13}$		0.32+.08-.07	2.66+.33-.29	1.35 (65)	$1.05 \times 10^{-13}$
81	0.95+.05-.13	0.65+.17-.06	1.54 (66)	$0.89 \times 10^{-13}$		0.41+.12-.09	3.15+.52-.40	0.95 (66)	$1.19 \times 10^{-13}$
84	0.15+.05-.05	2.92+.74-.52	1.20 (90)	$1.13 \times 10^{-13}$		0.35+.09-.07	2.62+.29-.25	1.21 (90)	$1.14 \times 10^{-13}$
89	1.04+.16-.17	1.01+.11-.27	1.51 (45)	$1.26 \times 10^{-13}$		0.42+.13-.10	2.80+.23-.36	0.75 (45)	$1.58 \times 10^{-13}$
94	0.14+.11-.09	1.78+.64-.34	1.07 (36)	$0.67 \times 10^{-13}$		0.72+.91-.33	4.87+2.56-1.72	1.24 (36)	$0.55 \times 10^{-13}$
99	0.29+.11-.08	3.59+1.07-.65	1.08 (48)	$2.22 \times 10^{-13}$		0.43+.13-.11	2.22+.28-.25	1.19 (48)	$2.22 \times 10^{-13}$
115	0.18+.05-.04	3.49+.79-.64	0.91 (106)	$1.71 \times 10^{-13}$		0.40+.10-.05	2.65+.34-.29	0.79 (106)	$1.67 \times 10^{-13}$

We fitted the spectra using either an absorbed one-temperature mekal thermal plasma model (Mewe et al. 1985; Kaastra 1992) or an absorbed power law model. The best fitting model parameters are listed in Table 4. In several cases, the fits with a power law with a rather large photon index $\Gamma$ are slightly better than those with the thermal plasma model. However, these models yield systematically larger $N_{\rm H}$ than the value $0.17 \times 10^{22}$ cm^-2 expected from the mean E(B - V) color excess (see Sect. 3 below). More complex models, such as two-temperature mekal or mekal + power law models could also improve the $\chi_{\nu}^2$. For instance, the spectrum of source #76 is better fitted ( $\chi_{\nu}^2 = 0.99$) with a 2-T thermal model with kT₁ = 0.84 and kT₂ = 5.55 keV. However, we caution that the quality of our spectra is usually not sufficient to distinguish a very hot ($kT \sim$ a few keV) multi-temperature thermal plasma emission from a power law spectrum.

For the 1-T mekal fits, we are usually left with a number of nearly equivalent solutions corresponding to either intrinsically harder, but less absorbed or more absorbed but intrinsically softer plasma models. Apart from the late O-star HD 164816 (source #19), all the sources have rather hard spectra with $kT \geq 0.65$ keV.

The spectrum of HD 164816 could be better fitted by adding a second (harder) component to the model, but the properties of this second component are not well constrained by the data. We note that the best fit model yields $N_{\rm H} = 0.43 \times 10^{22}$ cm^-2 which is significantly larger than the value of the interstellar column density $0.15^{+.05}_{-.04} \times 10^{22}$ cm^-2 derived by Diplas & Savage (1994) from the interstellar Ly$\alpha$ line. This suggests that part of the fitted absorption could actually arise in circumstellar (i.e. stellar wind) material. Correcting the X-ray flux for the pure interstellar column density only, we infer an "intrinsic'' (i.e. at the top of the stellar wind) X-ray luminosity of $5.7 \times 10^{31}$ erg s^-1. Adopting a bolometric correction of -3.00, we infer $\log{L_{\rm X}/L_{\rm bol}} = -6.96$ which is in pretty good agreement with the empirical $L_{\rm X} - L_{\rm bol}$ relation for O-stars proposed by Berghöfer et al. (1997). It is worth pointing out that Howarth et al. (1997) suggested that this star could be an SB2 binary (although no orbital solution exists so far). Provided that the interstellar column alone properly accounts for the low energy absorption, we find that the $L_{\rm X}/L_{\rm bol}$ ratio yields no evidence for an excess X-ray emission attributable to a wind interaction in a colliding wind binary.

If we assume that $N_{\rm H} = N_{\rm H, ISM} = 0.17 \times 10^{22}$ cm^-2 for the other sources, we find that most fits yield kT of a few keV. Note that the spectrum of SCB 731 (Fig. 6) corresponds to the X-ray emission during the flare event, the contribution of the quiescent time intervals to the spectrum being essentially negligible.

Figure 6: EPIC-MOS2 spectrum of the X-ray source # 66 associated with SCB 731. The solid line yields the best fitting absorbed mekal model.

Figure 7: Top: colour-magnitude diagram of the X-ray sources in the very young open cluster NGC 6530. The reddening vector with RV = 3.1 is indicated and the solid line shows the ZAMS relation taken from Schmidt-Kaler (1982) with a distance modulus DM = 11.25 and reddened with E(B-V) = 0.30 (see text). Black dots stand for stars from Sung et al. (2000) without H$\alpha$ emission, asterisks stand for pre-main sequence stars or PMS candidates from Sung et al. and the (overlapping) open diamonds indicate two stars classified as Be stars by van den Ancker et al. (1997). Bottom: unabsorbed X-ray luminosity (in the energy range 0.2-5.0 keV) as a function of the observed V magnitude. The MOS count rates were converted into luminosities assuming a 2 keV thermal plasma model, a uniform interstellar column density of $N_{\rm H} = 0.17 \times 10^{22}$ cm-2 and a distance of 1.8 kpc, except for the sources for which a detailed spectral fit is available. For these sources (symbols surrounded by an open circle) we used the flux from the actual fit. The luminosity of 9 Sgr is taken from Rauw et al. (2002). The dashed lines yield different values of $\log{(L_{\rm X}/L_{\rm bol})}$ (indicated by the labels) corresponding to main sequence stars of magnitude V in NGC 6530.