Astron. Astrophys. 361, 491-499 (2000)

5. The HI extent of NGC 3109

Huchtmeier et al. (1980, hereafter HSM80) measured the HI distribution across NGC 3109 with the Effelsberg radio telescope and a beam size of 9´. We apply the procedure used for the X-ray detections in the field of the SMC to the X-ray detections in the field of NGC 3109. First we make use of the HI map of HSM80 which has an extent of 60´ 30´. We convert the HI intensity into hydrogen column densities with the equation given in Dickey & Lockman (1990), cf. Kahabka (1999) and we derive a peak of . HSM80 have found that NGC 3109 has a large extent in HI , with an extension (distortion) of the HI in the SW. With our X-ray catalog we cover the whole extent of the HI of NGC 3109. If we restrict the analysis to the inner 20´ of this field and to sources with well constrained hardness ratios 0.2 then we can classify 7 sources as AGN.

A much higher resolution HI map of NGC 3109 with a beam size of 40" has been derived by Jobin & Carignan (1990, hereafter JC90) with the VLA. We now make use of this high-resolution HI image which we take from plate 67 of JC90. The HI distribution of NGC 3109 has an extent of 40´ 12´. The peak hydrogen column density is , and the lowest column density is .

In Fig. 3 we show the positions of the cataloged ROSAT PSPC sources overlaid on the gray scale HI image of NGC 3109 and taken from JC90. We find 26 ROSAT PSPC sources within the HI contours of JC90. If we restrict the analysis to the inner 20´ of this field and to sources with well constrained hardness ratios 0.2, then we can classify 3 sources (number 36, 41 and 63) as AGN and 2 sources (number 53 and 73) as X-ray binaries. Source 53 may be seen through higher gas columns while source 73 is seen through lower gas columns (see Fig. 4 and Eq. 11). One source (number 82) can be either class. The two sources 52 and 92 cannot be classified as AGN or X-ray binaries. Source 92 may be a foreground object (cf. Fig. 4, Fig. 5 and Table 2). Source 52 has similar hardness ratios as the LMC SNR 0548-70.4 (Haberl & Pietsch 1999). It could be a (young) SNR in NGC 3109.

Fig. 3. Positions of ROSAT PSPC X-ray sources (the labeled circles) overlaid on an HI image of NGC 3109 obtained with the VLA (JC90) as shown in gray scale. The maximum hydrogen column density is and the lowest hydrogen column density is .

Fig. 4. Hydrogen absorbing column density - hardness ratio (upper panel) and (lower panel) plane for 10 X-ray sources within 20´ of NGC 3109 with well determined (0.24) and (0.24). The hydrogen column density is the NGC 3109 intrinsic and galactic foreground value. For NGC 3109 the HI model of JC90 is used. Also given are the theoretical curves for X-ray binaries (powerlaw slope -0.8) and AGN (powerlaw slope -2.0 to slope -2.6) assimung a metallicity 0.2 solar.

Fig. 5. - plane for 6 ROSAT sources projected onto the HI extent of NGC 3109 and for which accurate values and have been determined. Also shown are the two AGN tracks for powerlaw photon index -2.0 (upper track) and -2.6 (lower track) and the X-ray binary track (dashed line).

Table 2. The 9 classified sources in the central 20´ of NGC 3109 using the HI data of JC90. A = background AGN, B = X-ray binary source, F foreground star, AB = class A or class B.

NGC 3109 has a mass smaller than, or comparable to, the LMC and 2 X-ray binaries with luminosities above a few times would be in agreement with extrapolations from ROSAT findings for the LMC (cf. Haberl & Pietsch 1999).

There are 14 ROSAT PSPC sources which are projected onto NGC 3109 intrinsic hydrogen columns of (cf. Fig. 3). The largest column is derived for source 86 (). This source could be associated with a spiral arm or an HII region of NGC 3109. Two further sources, 53 and 59, are close to another region of large column density and source 59 is also close to the optical center of NGC 3109.

We further test the correctness of the classification by constructing the distribution of the number of detected X-ray sources N with fluxes in excess of a given flux S, the (cf. Hasinger et al. 1993). We restrict the analysis to sources within the central 20´ of NGC 3109 which have been classified as AGN or X-ray binaries. We correct the X-ray fluxes for the intervening hydrogen columns by using for the galactic contribution a value of (Dickey & Lockman 1990) and for the contribution due to NGC 3109 the HI model of JC90. We find that the observed is in excess of the of the soft extragalactic X-ray background (Hasinger et al. 1993). This can be accounted for if an additional less steep component (e.g. due to X-ray binaries) is added. If we only construct the of the candidate AGN, then the of the soft extragalactic X-ray background is reproduced. This would mean that no significant additional hydrogen is needed to explain the observed sources as background AGN and X-ray binaries although only a few candidate AGN are within the HI extent of NGC 3109. The two brightest candidate AGN (source 36 and 63) could give an excess in the . We cannot exclude that they are at least in part related to NGC 3109.

© European Southern Observatory (ESO) 2000

Online publication: October 2, 2000
helpdesk.link@springer.de