Previous studies of the line-profile variations in HD 200775 concentrated on their short-term behaviour (e.g. Ruusalepp 1987, Beskrovnaya et al. 1994) or did not make extensive use of radial velocity information (e.g. Miroshnichenko et al. 1998). In the present study we consider the long-term behaviour of the line profiles, equivalent widths, and radial velocities, and therefore we have a sounder basis for discussion of the possible causes of the observed phenomena.
In the 1994/5 low-state phase HD 200775 displayed broad double-peaked emission in the H and H lines (see Fig. 1). Profiles of this type are usually observed in classical Be stars, which are commonly recognized to be surrounded by rotating gaseous disks (see Waters & Marlborough 1994 for a review). Doppler broadening owing to the rotation of the CS disk is probably responsible for the great width of the Balmer emission lines. It can also explain the width of the absorption wings of He I 5876 (see Fig. 2) if the line-forming regions extend close enough to the star to be seen in projection against the disk at high radial velocities. The results of radio observations by Lepine & Rieu (1974) and Watt et al. (1986) confirmed the existence of a large, possibly relic, accretion disk near HD 200775 which is seen as a flattened molecular condensation along with two opposite outflowing streams. Thus, the line-emitting, gaseous disk near the star may be the inner part of a global CS disk.
Axial rotation was not the only regular large scale motion in the gaseous envelope of HD 200775 in 1994-1995. The presence of redshifted central absorptions in H and H indicates that matter infall from the envelope onto the star also took place during this interval, as well as during the more active phases.
Also observed before the high-activity phase was a variable blueshifted emission bump in the He I line (Fig. 2), which we have interpreted as a signature of the non-stable wind. Numerous indications of the stellar wind from HD 200775 were noted earlier by a number of authors (see Beskrovnaya et al. 1994for a review). The co-existence of CS components with opposite kinematics, even in a low-state phase, suggests that the gaseous envelope around HD 200775 is likely to be latitudinally stratified, with the wind region being situated at higher latitudes, outside the equatorial accretion disk. A similar geometrical model, which contains a dense equatorial disk and a fast, but less dense, wind close to the polar region was suggested for classical Be stars in the early 1980's (Poeckert 1982, scenario 2). A similar spatial distribution of the wind and accretion in the CS gas for the whole class of Herbig Ae/Be stars has been suggested by Grinin & Rostopchina (1996) on the basis of a statistical study of their H line profiles.
A clear signature of the outflow appeared in both H and H lines during the high-activity phase. The rise of emission in these lines implies an increase in the emitting gas mass, which, in turn, is most likely due to an additional generation of the stellar wind. Thus, both the wind and the accretion controlled the global radial motion in the envelope at that time. In order to understand the origin of the wind, it is necessary to interpret possible connections between different events taking place during the whole high-activity phase, whose evolution is illustrated in Fig. 3. Two alternative phenomena can be considered as a possible primary factor stimulating the growth of the H and H emission: the appearance of a strong wind, displayed in the blueshifted absorption components, and the monotonic velocity increase of the accreted matter, indicated by the redshifted feature. For the time being, we cannot determine whether the behaviour of the He I line profile is causally connected with the variability of H and H. Since the new increase of the He I core velocity in 1999 was not followed by corresponding changes in H and H, the observed correlation between the velocities of the absorption core and of the blueshifted H and H absorptions in 1997 may have been accidental. Further investigation is needed to clearify this question.
As one can see in Fig. 3, the wind signatures in H and H were strictly correlated with the EW(H) variability, whereas the increase of the accretion velocity was seen about five months after the EW passed through its maximum and started fading. This favours the following hypothesis: a) it was the strong wind that was responsible for the active phenomena observed in 1997, and b) the wind in HD 200775 is not powered by disk accretion, as was suggested by Grinin & Rostopchina (1996), but is an independent factor connected more with the star itself than with the CS disk (Böhm & Catala 1995, Pogodin 2000).
Thus, the results of our study support the scenario of an interaction between the stellar wind originating from the star and the rest of the CS envelope (disk) proposed by Beskrovnaya et al. (1994) for interpretation of the phenomena observed during a previous high-activity phase of HD 200775 in 1986. No rigorous theory of the wind generation from the surface of both classical Be and Herbig Ae/Be stars has been developed yet. However, some possible mechanisms which may give rise to the wind, based on the stellar magnetic activity or a wind/disk interaction, have been proposed by Smith (1989), Hanuschik et al. (1993), Bjorkman & Wood (1995), and Strafella et al. (1998). In particular, according to Hanuschik et al. (1993), the H emission rise might be caused by matter outflow from the star due to magnetic flares. A certain fraction of the ejected material is expected to attain circularized orbits due to collisions and to form thereby a CS near-Keplerian decretion disk. Interaction between the basic accreting disk and the decreting one, which can exist some time more after the end of the wind process, could stimulate additional matter infall, similar to that observed in HD 200775 during five months after the end of the main wind event (Fig. 3, middle panel). In the 1986 high-activity phase this infall did not have a stable (as in 1997), but rather a discrete, character (Beskrovnaya et al. 1994).
A completely alternative interpretation of the observed phenomena can be suggested if one assumes that HD 200775 is a binary system. This is a relevant suggestion given that cyclic events have been detected. The following studies have been done in an attempt to check whether the object is a binary. The results of near-IR speckle interferometry (Millan-Gabet et al. 1999) suggested the presence of a secondary component at a distance of 20 milliarcsec from HD 200775. However, no information about the secondary has been determined from these data. A search for periodicity in the radial velocities of several lines in the spectrum of HD 200775 was carried out by Corporon & Lagrange (1999), but their data turned out to be insufficient to draw any certain conclusions. Nevertheless, the question of the object's binarity remains open and awaits new high-quality data.
We are planning to continue spectroscopic monitoring of this interesting object in order to address the following problems: a) binarity of the object; b) cyclicity of its stellar and CS activity; c) construction of a detailed picture of the phenomena observed in HD 200775 at different phases of its activity. We would like to emphasize that the cycle period is still uncertain, since the detailed observations have been obtained only around one high-activity phase. For example, the redshift of the HeI line observed in 1999, the reasons for which are not understood yet, suggests that a source of activity near the star still exists, while the H line passes through its low-state. In this light, follow up observations during 2000-2002 and especially around the next projected maximum in Spring 2001 are of high importance.
© European Southern Observatory (ESO) 2000
Online publication: June 30, 2000