4. Discussion and conclusions
There are at least three physical mechanisms which could account for the light variability of the TTS (see e.g. Herbst et al. 1994).
The first mechanism is an intrinsic one - the magnetic activity manifested in starspots resulting in (periodical) rotational modulation of the stellar brightness (the typical example is V410 Tau).
The second is related to interaction of the star with its circumstellar environment - irregular accretion of matter onto the star. This mechanism is probably dominant in the classical TTS, where accretion components are occasionally observed in the line profiles.
The third is an extrinsic one - obscuration of the star by circumstellar dust clouds. This phenomenon is most clearly observed in the UX Ori-type stars (Grinin et al. 1994), which are earlier spectral type counterparts of the TTS. It was also observed in the TTS RY Lup (Gahm et al. 1989). In a classical TTS, all three mechanisms may be in operation, which makes the observed variability very complicated.
The event of brightening of RY Tau, discussed in this paper, constitutes probably the rare case (for a classical TTS), when only one mechanism is dominant, while the others were not as efficient. Although the inflow of gas is evident through the red-shifted components of the sodium doublet lines and the He i line, the absence of measurable veiling indicates that accretion processes do not affect the brightness of the star. Neither is the presence of dark spots evident: no significant periodicity was found in the light variations. The case of a large polar spot of variable effective area can also be excluded: such a spot would be visible in the photospheric line profiles.
The constancy of the photospheric parameters indicates that the source of the light variability is extrinsic - the dusty circumstellar environment has changed its opacity along the line of sight, which was observed as an apparent brightening of the star. The increase of the linear polarization accompanying the decrease in brightness supports this conclusion. The circumstellar dust is most probably confined to a disk, i.e. we observe RY Tau almost equator-on. The large value of also supports this assumption.
The fact, that the flux radiated in the H emission does not correlate with the stellar brightness, may indicate that the emission line source is located mostly outside of the star. When the star is obscured by dust, the contribution from the emission line source becomes larger and the emission spectrum appears more prominent, while the flux radiated in emission lines remains about the same or varies independently.
The strengthening of emission lines at minimum brightness was reported earlier by other observers. From simultaneous photometric and spectroscopic observations, Holtzman et al. (1986) found that H changes from absorption, when the star is bright (V), to emission, when the star is faint (V), and the forbidden line [O i] 6300 Å became visible when the star was faint. Analysis of IUE spectra showed that the line Mg ii 2800 Å changes from absorption to emission when RY Tau is fading from V to V (Eaton & Herbst 1995). It was noticed also already by Herbig (1961) that the forbidden line of [S ii] was more prominent in emission when the star was faint.
From all this, we may conclude that the emitting region is not screened by the circumstellar dusty disk, although we look at the star almost equator-on. The result is quite expected for the forbidden lines, which form in a region extending far out of the star. It could also be expected for H, which comes from a large volume of stellar wind, but was not expected for the Ca ii emission, which is often treated as an indicator of chromospheric activity. The appearance of emissions in the Mg i and Fe ii lines at lower brightness is also surprising. We may conclude that at least H and the Ca ii emission lines in RY Tau are formed neither in a chromosphere nor in an atmosphere of a boundary layer, otherwise they would be subject to obscuration by dust as well as the central star.
© European Southern Observatory (ESO) 1999
Online publication: December 4, 1998