Astron. Astrophys. 318, 73-80 (1997)

5. Discussion

5.1. The extended spray region at the disk rim

As a consequence of the extended spray model the white dwarf is never directly observable. This is supported by the observation of a rather low X-ray luminosity - erg/sec (Pakull et al. 1988, Greiner 1995, private communication) compared to other known X-ray binaries with luminosities on the order of erg/sec.

Simulations of relatively thin hot spot regions which allow a view to the white dwarf at each phase did not fit the data. Their luminosity is to small by two magnitudes. Additionally, the direct view to the white dwarf described as a point source would result in a deep and wide eclipse which would last about 0.12 in orbital phase. Another problem with thin disks is the depth of the secondary minimum which is caused by the disk covering a part of the star (see models a and c). An extended spray, on the other hand, generates the observed depth.

An interesting point to note is the scattering of the data of Schmidtke et al. (1993). Apparently it is less during the eclipse, but this is misleading: Subtracting a mean light curve one obtains a constant amplitude of the deviations, even during eclipse, see Fig. 6. Therefore, the scattering is produced in a region larger than the disk/star system. Because of the observed P Cygni profiles in some systems a wind is a promising candidate for the scattering medium. This would also explain the less deep X-ray eclipse.

Fig. 6. The observed data (Schmidtke et al. 1993) minus the mean light curve. The scattering 36 223 576 568does not depend on the orbital phase suggesting a scattering region much larger than the disk size.

5.2. Comparison to other supersoft sources

There are at least two other supersoft sources with well observed optical light curves, Cal 83 (Smale et al. 1988) and RX J0019 (Beuermann et al. 1995). Both light curves show no eclipse indicating a lower inclination than that of CAL 87. This raises the question how the light curve of CAL 87 would appear if we saw the system under other inclinations. We show these light curves in Fig. 7 and superimposed the observations of Cal 83 (Smale et al. 1988) and RX J0019 (Beuermann et al. 1995). We find good agreement between the simulation of CAL 87 seen under and Cal 83, and also between the simulation and RX J0019. The latter is quite different from the value estimated by Beuermann et al. (1995) from Doppler shifts of HeII lines in the case that they originate near the white dwarf. A recent investigation of phase resolved UV data (Gänsicke et al. 1996) as well as recent photometry of RX J0019 (Will & Barwig, 1996) point to a higher inclination as indicated by the secondary minimum. Simulating different inclinations of model b shows, that in the case of no energy transport the inclination should be even a bit higher to reproduce the observed orbital light variation.

Fig. 7a and b. The light curves of CAL 87, model d, if the system would be seen under different inclinations. Both diagrams show solutions for inclinations (top) to (bottom) in steps of . Additionally we scanned the light curves of RX J0019 (upper panel, from Beuermann et al. 1995) and CAL 83 (lower panel, from Smale et al. 1988). For CAL 83 Smale et al. plotted also the best-fitting sine wave.

The good agreement suggests common main features in SSSs. This supports models where the spray dominates the optical light curves, because the unusually high accretion rate resulting in a strong interaction of stream and disk rim is common to all of them. We have not computed light curves for other system parameters as higher masses or larger disks. The periods of the binary seem to play a minor role in this context. This is not surprising if we keep in mind, that the geometry of CAL 87 depends only on the mass ratio of the binary stars and that the possible mass ratios according to the vdH model lie in a narrow range.

© European Southern Observatory (ESO) 1997

Online publication: July 8, 1998
helpdesk.link@springer.de