5 Discussion

5.1 System's parameters

FS Aur is a non-eclipsing binary, so information on the system's parameters is difficult to obtain accurately. Usually for definition of the parameters of such systems some empirical formulae are used. However, it should be noted that these assessments are subject to unknown and potentially large errors and should be adopted with appropriate caution. In the present paper we have decided to restrict ourselves only to estimation of the parameters. To ensure some validation, we have used only stringent relations. Thus we believe that the restrictions on the basic system parameters for FS Aur obtained by us are correct, if the values of input data are correct too.

In this connection we pay attention to the obtained estimate of the mass of the white dwarf which is quite small. We have estimated the primary mass to be less (and possibly much less) than 0.46 $M_{\odot}$, which is near to the lower limit of the observed range for white dwarf masses in cataclysmic variables (Webbink 1990; Sion 1999). The estimate of M₁ depends strongly on the estimated K₁ value. Let us discuss possible errors in the definition of K₁.

The definition of the semi-amplitude of the radial velocities of the emission lines, really reflecting the orbital moving of a white dwarf, is a very complicated problem. The contribution to a broad emission line can be introduced by many emission areas of a binary system. For example, Balmer emission from the secondary star forms an additional component in an emission line moving with semi-amplitude K₁/q. This component upon condition of small q can deform the line wings. Any nonhomogeneity of the accretion disk can cause even greater distortion in the line profile.

In our case the contribution of the secondary to the Doppler maps is completely absent. At the same time, the Balmer and Helium emission is distributed very nonuniformly (Figs. 9 and 10). In general, this factor might affect the accuracy of the definition of K₁, but we hope that it has not occurred. Actually, though the areas of bright inhomogeneities can be detected rather clearly on the Balmer tomograms, nevertheless they reach distances of no more than about 500 kms^-1 from the center of the tomograms. We would like to recall, that we tested the line profiles at 650-750 km s^-1 from its center where inhomogeneities became less noticeable.

On the other hand it is necessary to note that TPST have found a somewhat smaller radial velocity semi-amplitude ( K₁ = 60 kms^-1). However they have noted that the main aim of their research was the definition of the orbital period, and their values of K₁ should not be used in dynamical solutions of the system. When we determined K₁ we adopted values of $\Delta$ much smaller than FWZI. The dependence of parameter $\sigma(K)/K$ on the Gaussian separation for H$\beta$and H$\gamma$ lines is very similar (Fig. 6). It can be seen that $\sigma(K)/K$ decreases monotonically with increasing $\Delta$. Having selected a greater value of $\Delta$ we can of course obtain a smaller value of K₁. However we cannot offer any convincing reasoning for increasing $\Delta$. It even seems undesirable to do so, as increasing $\Delta$ will lead us into the outer line wings, which are subject to contortion at some phases (Figs. 2 and 3).

Finally, it is necessary to pay attention to the emission ring, which is well noticeable on the H$\gamma$ and HeII tomograms. It is tempting to connect it with the accretion disk, and the bright spot in the ring center with the white dwarf. In this case it would become possible to independently determine K₁. Unfortunately, such a ring corresponds to a too large accretion disk, which cannot lie in the Roche lobe of the white dwarf.

Thus we believe and hope that the value of semi-amplitude of K₁(and system parameters) obtained by us is correct. Nevertheless we consider, that new, longer and better-quality observations are extremely necessary for a more precise definition of the system parameters of FS Aur.

5.2 The structure of the accretion disk

Another important result of this work is the spot structure detected in the accretion disk of FS Aur. Doppler tomography has shown at least two additional bright regions in this system. The first, brighter spot is located at phase about 0.6. The second spot is located opposite the first and occupies an extensive area at phases about 0.85-1.15. The detected spot structure of the accretion disk is confirmed by a dependence of equivalent widths on orbital phase (Fig. 8). The observed minima of EW can be due to an increase in the continuum luminosity when the enhanced emission region crosses the line-of-sight.

An enhancement of the emission coming from the back of the accretion disks of some cataclysmic variables was noted by many observers (see review by Livio 1993). Some theoretical and numerical studies indicate that high, free flowing gas could pass over the white dwarf and hit the back side of the disk (Lubow & Shu 1976; Lubow 1989; Kunze et al. 2001). In this case, the azimuth angle of the region of "secondary interaction" should be about $140^{\circ}$- $150^{\circ}$ for a wide range of the system's parameters, while the distance from the accreting component will change from 0.02 to 0.18 of the system's size, depending on the mass ratio (Lubow 1989). The phase where the line-of-sight crosses this bright spot must be about 0.6. This is where our observed bright spot is found!

In addition, recent numerical hydrodynamic calculations point to a possible difference in the thickness of the outer edge of the accretion disk in a close binary system. For example, Meglicki et al. (1993) identified three thicker regions of the disk at phases 0.2, 0.5, and 0.8. Armitage & Livio (1996) also pointed to a possible transfer of the stream matter above the plane of the accretion disk and an increase in the number of atoms along the line-of-sight relative to the average level at phases 0.1-0.2 and 0.7-1.0. Thus, the second bright region of FS Aur can be connected with one of the accretion disk's thickening regions found by Meglicki et al. (1993). The mechanism of the increase of its luminosity is not quite clear, but it is probably attributable to ionization by emission from the inner disk's region.

The nature of the ring-shaped structure visible on H$\gamma$ and HeII tomograms remains completely unintelligible. As was already noted above, this structure cannot be connected with the accretion disk, as its size should be so large, that it cannot lie in the Roche lobe of the white dwarf. Another possible site of origin is a nebula. However, the radial velocity of the nebula's center should coincide with the systemic velocity of the binary. In our case, though, the velocities are close, but still noticeably different. We again come to the conclusion, that new and better quality data are needed for the explanation of this puzzle.