Astron. Astrophys. 318, 819-834 (1997)

8. Conclusions

Time series of high-resolution optical spectroscopic observations of three early-B hypergiants have shown that their winds are highly variable.

The "photospheric" lines show cycles of pulsation-like radial-velocity variations, with a period of about 12 days. These variations last a few cycles before becoming irregular. By integrating these variations we expect the radius variations to be of the order of 10% of the stellar radius. These radius variations are connected to the emission variability. The photometric data indicate a possible radial pulsation with this amplitude.

The mean photospheric lines of metals and the mean profiles of the Balmer emission lines have been used to derive stellar parameters and wind parameters. The terminal velocities of the winds are between 370 and 515 . Presumably, the stars are on the "low side" of the bi-stability limit of stellar winds (Lamers et al. 1995). This is in agreement with the low degree of ionization in the wind. The mass-loss rates derived from the Balmer emission lines are between 2.7 and , in very good agreement with those derived from radio data. The velocity law in the stationary wind has a value of =1.5 to 1.7.

The variations of the photospheric lines and the wind lines have been studied in detail. The variable P Cygni-type profiles of the wind-sensitive lines are particularly distinguished by propagating discrete absorption components with a typical repetition time of 24 days.

These discrete absorption components can be understood as non-spherical density perturbations, i.e. "blobs". These absorption components are already present at very low velocities. This implies that they do not originate in the wind but are triggered by perturbations in the photosphere or even deeper layers of the star. The propagation of such components can be traced from nearly photospheric velocities to times the terminal velocity.

This behavior is very similar to the well-known DACs in O stars, which are present in the UV as well as in the optical range. These DACs also have slower parameters than the ambient wind and are explained as propagating denser regions. However, O stars have much higher terminal velocities and much shorter flow times through the line-forming region in the wind. Also the time scales for stellar variability are shorter.

In stars with broad wind emission lines like Wolf-Rayet (WR) stars, variable structures seen at the top of the emission lines are observed (e.g. Moffat & Robert 1991). These emission subpeaks are ascribed to blobs in the wind. Such subpeaks are not observed in early-B hypergiants. One possible explanation is that the blobs in hypergiants are less "discrete" than in WR stars. The observability of such substructures depends on the density contrast between the ambient wind and the blobs. In our stars this is in the order of 15%.

Also, in early-B hypergiants, the width of the observed absorption features in time is of the same order as the wind-flow time. The width of the emission component is much smaller than in a WR star. Thus we expect that the "blobs" affect the whole width of the emission profile.

By comparing our data with theoretical -type velocity laws we find significant deviations in the low-velocity regions of the wind. The features have nearly constant acceleration.

The innermost zone of acceleration is observable in the variations of the photospheric lines. The velocity increases much faster than predicted by a =1.5 velocity law which was derived from the stationary wind model.

The wind activity of early-B hypergiants differs remarkably from the flow characteristics of late-B/early-A supergiants, which have been investigated recently in detail by Kaufer et al. (1996a, b). In these objects, the wind is primarily modulated by a mechanism locked to photospheric rotation due to distinct surface features, presumably magnetic surface structures. These pronounced differences in stellar-wind properties clearly demonstrate the need for spectroscopic long-term monitoring campaigns. Such campaigns provide important input data for a better understanding of the various physical processes responsible for the flow variations in luminous hot stars and the wind activity in a broad range of stellar types.

© European Southern Observatory (ESO) 1997

Online publication: July 3, 1998
helpdesk.link@springer.de