Astron. Astrophys. 363, 1029-1039 (2000)

5. Discussion

5.1. Phase timings

Fig. 11 and Fig. 13 show the orbital trailed spectra of H and HeI folded with the orbital period along with the corresponding Doppler reconstructions (see below) ⁴. The phase used here is consistent with the photometric phase expected under the assumption that the photometric phase reflects the orbital phase (i.e., inferior conjunction of the secondary star occurs at ). The spectra of corresponding to the rotational disturbance, whose phasing in the trailed spectra of the individual nights seem to be slightly different within about 0.2 phase were removed prior to mapping, because they possibly affect the results. Each reconstructed spectrum therefore shows voids at the corresponding phases.

The trailed spectra show both of the double peaks of the emission lines and a clear S-wave modulating independently the global emission pattern with the orbital frequency. These features are the main cause of the orbital variability of the radial velocity corresponded to the spectroscopic ephemeris (Eq. 2). Although weaker, the same phenomenon is also seen in the HeI emission line (Fig. 13).

The fact that the phases of the rotational disturbance (), the zero-crossing phase of the H wings (), and the photometric eclipses () agree within the errors gives us some confidence that we know the orbital phasing of the binary components (Fig. 5, Fig. 8).

5.2. System parameters

The mass function of the binary system can be derived from the semi-amplitude of radial velocities of the white dwarf and the orbital period :

where and G is the gravitational constant. The mass of the secondary star can be independently estimated from the empirical relation between and for main-sequence secondaries of CVs for (Warner 1995):

The mass function and determine a relation between a mass of the primary star and the inclination angle of the binary system (Fig. 10). The relation immediately indicates that should be less than within the errors of for the object. If is adopted as a reasonable inclination angle for a system with shallow eclipses (e.g., Fig. 4 of Bruch et al. 2000), we obtain 0.5-0.6 and then a mass ratio 0.2-0.3.

Fig. 10. Expected distributions of the white dwarf mass (, in solar masses) with inclination angle of the binary system assuming and .

Although no superoutburst has been detected so far for the object, one may anticipate that the object is a member of the SU UMa-type dwarf novae and will show superoutbursts sometime, based on the orbital period just below the period gap. Superoutbursts are characterized by superimposed superhumps during the bright state. Theoretically, superhumps are expected to be caused by a tidal instability of an eccentric accretion disk of the SU UMa stars by an enhanced 3:1 resonance of a rotational velocity at the outer radius of the accretion disk and the orbital motion (Osaki 1989). A mass ratio is required for the tidal instability (Whitehurst 1988; Hirose & Osaki 1990; Whitehurst & King 1991). If the expectation of the white dwarf is correct, the object should satisfy the criterion.

5.3. Time-resolved Doppler maps

Fig. 12 and Fig. 14 present the two-dimensional Doppler maps of the H and HeI emission lines; the fit converged to a reduced of 1.6. A mass accretion stream and a lobe of the secondary star are drawn only for the map of April 18 (top-left) using the and the corresponding .

Fig. 11. Observed and reconstructed trailed spectra of H for the individual nights (April 18 and 19 (top), April 20 and May 29 (middle), May 30 and 31 (bottom) for left to right, respectively). For each night, two images of the observed (left) and reconstructed (right) spectra are arranged side by side. The vertical axis corresponds to the photometric phase. The rotational disturbance is removed for the reconstructed spectra, i.e., spectra of the phase 0.1-0.8 were used for the mapping (-0.9 were also excluded as a preventive measure). The horizontal axis represents the velocity ( cm s-1).

Fig. 12. Doppler maps of H (April 18 and 19 (top), April 20 and May 29 (middle), May 30 and 31 (bottom) for left to right, respectively), that is made removing the phase corresponding to the rotational disturbance. The horizontal and vertical axes represent the velocity (cm s-1).

Fig. 13. Trailed spectra of HeI 5876. The arrangement of the figures is same as that of Fig. 11.

Fig. 14. Doppler maps of HeI 5876. The arrangement of the figures is same as that of Fig. 12.

The dominant features observed in the H maps are two isolated blobs on top of the weaker ring-like accretion disk component. The left blob is almost stationary in the H and helium maps, indicating that it is the classical hot spot. The right blob, always present in H and definitively absent in the helium map, probably arises from the irradiated secondary star.

This emission disappeared on May 31, which indicates that the H emission from the irradiated secondary star decreases during the pre-outburst stage. Models of irradiated secondary stars of short period dwarf novae, show the incident flux produces marked changes in the temperature stratification and emitted spectra of the irradiated star (Brett & Smith 1993). Increased irradiation warms the upper layers of the chromosphere, decreasing the temperature gradient. This effect might induce the Balmer lines to change rapidly from emission to strong absorption during outburst (Houdebine et al. 1995).

Even more remarkable are the Doppler maps of the HeI line; the H and HeI 5876 maps show different features. While a significant fraction of the H emission comes from low velocity regions, suggesting an origin near the outer accretion disk, the bulk of the helium emission arises from regions of higher velocity, and the outer edge of the accretion disk is invisible in the helium maps, perhaps caused by self absorption of the emission line. On May 31, the HeI emission is dominated by four distinct emission regions superimposed upon a ring of emission at , all located on the trailing edge of the disk.

The component of the irradiated secondary star on the Doppler map of H provides us the radial velocity semi-amplitude of the secondary star . Since the emitting region likely represents the irradiated side of the surface, we adopt approximately at the center of mass of the secondary star. Then, we obtain and . This result is consistent with Fig. 10 within the errors and confirms the estimation of the system parameters given in the previous subsection, indicating that we can observe V893 Sco as a potential SU UMa-type dwarf nova even if there still is no detection of superoutburst.

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk@link.springer.de