Astron. Astrophys. 344, 459-471 (1999)

5. The spatial distribution

The 26 supersoft sources found in M31 are distributed over the whole galaxy disk (cf. Fig. 4 for the distribution of the SW and the C-sample, cf. also Fig. 14 and Fig. 15). This may be in favor for a disk population. Recent spatial studies of novae in M31 using a Monte Carlo simulation have been performed by Hatano et al. (1997). From this study it follows that the ratio of bulge to disk population is about 1/2 similar to the ratio of bulge to disk mass of this galaxy although there is some controversy about this subject. In comparison for the Milky Way a ratio of 1/7 is found both for the novae and the mass. The M31 bulge has in the model of Hatano a radius of 6 kpc (equivalent to a projected size of 0.5^o). A more extended discussion on the distribution of novae in M31 can e.g. be found in Capaccioli et al. (1989), Rosino et al. (1989) and Yungelson et al. (1997). Interestingly there are 6 of the 31 (likely) candidate supersoft sources found in the bulge region. But three of these objects correlate eiher with a galactic foreground star or a M31 supernova remnant. A ratio 1/(5-10) is derived for supersoft sources detected in the bulge compared to objects detected in the disk. This ratio reduces to 1/(4-7) if only "accepted" systems are considered (cf. Fig. 15). There may be a chaining of supersoft sources (from the C-sample) along a (12-16) kpc arm (cf. Figs. 4, 5, 14 and 15). This feature may also be found in the model of Hatano (cf. his Fig. 3). The spatial distribution of the SW-sample (with a mean radius of 14 kpc) is consistent with a 12-16 kpc spiral arm of the M31 galaxy (cf. Fig. 6 of Braun 1991). Another grouping of supersoft sources in a 18-21 kpc ring (cf. Fig. 15) may be connected to a 18-24 kpc spiral arm (cf. Braun 1991). There are no detections within the 6-12 kpc ring (the only exception may be the source with the catalog index 212). In this ring the (projected) hydrogen column reaches values up to .

Fig. 14. Distribution of supersoft sources in M31 (solid histogram) as a function of the galactocentric distance in M31. Also shown is the radial distribution of the hydrogen column density for the NE Urwin model (dotted line) and for the SW Urwin model (dashed line) which is not corrected for the inclination of the galaxy. The catalog indices of individual sources are given (cf. Table 1).

Fig. 15. Upper panel: galactocentric distribution of supersoft sources in M31. Separate histograms are shown for objects not correlating with a foreground star or a M31 supernova remnant and located in the upper (n1), lower (n1) galaxy hemisphere. Also the distribution of "intrinsically absorbed" sources (n2) is shown and the distribution for both hemispheres (0n2). Lower panel: galactocentric distribution of novae (from Sharov & Alksnis 1991, 1992), M31 blue stars with B-V0.3 (from Magnier et al. 1992 and Haiman et al. 1994) and Cepheids (the Baade and Magnier sample, cf. Magnier et al. 1997).

In Fig. 15 we also show the galactocentric distribution of the M31 novae (from Sharov & Alksnis 1991, 1992), blue stars (B-V0.3) and Cepheids (from Magnier et al. 1992, 1997 and Haiman et al. 1994). Novae belong to an old stellar population and are preferentially detected in the bulge (at galactocentric radii 6 kpc). Blue stars belong to a young stellar population and trace the galaxy light. Cepheids belong to a somewhat older population. The distribution of Cepheids trace the spiral arms of M31 where recent star formation is taking place. They are found (in the distribution) predominantly within 8-15 kpc. For a consideration of the completeness of the Cepheid sample see Magnier et al. (1997). Actually there could be a second Cepheid peak at galactocentric radii 18-22 kpc where another M31 spiral arm is found and where a peak in the supersoft distribution is found. But the Magnier survey apparently did not cover this region.

While novae appear to be bulge-dominated in the observational sample most possibly due to the low dust content of the M31 bulge supersoft sources appear to be to a less degree bulge-dominated and are more likely associated with the spiral arms. Considering only objects with 0n2 then supersoft sources are found within galactocentric radii of 5-25 kpc. They are not found in the bulge of M31, but within the range of Cepheids and blue stars and at 18-22 kpc. Bulge sources at galactocentric radii r6 kpc are only found in the "intrinsically absorbed" sample. They may be consistent with classical or symbiotic novae as these objects show high intrinsic absorption and tend to belong to an old population. The M31 supersoft sources may belong to a younger population similar to the Cepheids but to an older population than blue stars. Hatano et al. assume a scale height of the M31 novae of 350 pc. We find that the scale height of the M31 supersoft sources is consistent to be smaller (100-150 pc, at 5 kpc) which favors a younger stellar population and is in agreement with the supported view that slightly evolved main-sequence stars or subgiants are involved (van den Heuvel et al. 1992). A mean space density can be inferred for the population of supersoft sources in M31 (assuming that they are homogeneously distributed in a disk of radius 20 kpc and have a scale height of 150 pc) of . One could suspect from Fig. 15 that we see two different sub-populations of supersoft sources, one in the bulge at galactocentric radii 6 kpc, possibly associated with (classical and symbiotic) novae, one at radii 12-16 kpc and 18-20 kpc respectively tracing spiral arms and possibly associated with subgiants and CV-type supersoft sources and one at radii 18-22 kpc also associated with a spiral arm and possibly associated with subgiants and CV-type supersoft sources.

© European Southern Observatory (ESO) 1999

Online publication: March 18, 1999
helpdesk.link@springer.de