1.3 The Distribution of Circumstellar Material

Circumstellar material is a characteristic of hot luminous stars (Dougherty1993); it is ubiquitous and often remains present throughout their entire life (Waters et al.2000). The stars’ spectra show complex hydrogen Balmer line profiles with emission components. In addition to Be stars, stars such as O and B supergiants, luminous blue variables, Wolf-Rayet stars Oe and Ae stars all exhibit circumstellar matter.

Be stars have two distinct regions of circumstellar matter: a diffuse polar stellar wind and a denser, Balmer line emitting, component in the equatorial plane of the central star (see e.g.,  Dachs1987Dachs et al.1986Porter1998Slettebak1988). The fast diffuse polar wind is well described by standard line-driven wind theory physics (see Castor et al.1975Lucy and Solomon1970), however, it has been much more difficult to describe the equatorial disc (Porter1998).


Maximm-entropy reconstruction of the H emission of the B shell star
Figure 1.1: Maximum-entropy reconstruction of the Ha emission of the Be shell star z Tauri. The arrow indicates the position angle of the linear polarization (Quirrenbach et al.1994).

The distribution of circumstellar material in Be stars giving rise to the emission features in their spectra is concentrated in the equatorial plane. Direct evidence of this highly flattened geometry of Be star envelopes has come from imaging at optical (e.g., Quirrenbach et al. 19971994Stee et al. 19951998) and radio (Dougherty and Taylor 1992) wavelengths. The Cambridge Optical Aperture Synthesis Telescope (COAST) team, have also been able to demonstrate direct imaging of Be star discs.

Optical interferometry is now able to provide milli-arc-second spatial resolution and can be thus used to probe the circumstellar environments of Be stars. Of note is Figure 1.1 which shows direct observational evidence of extended Ha emission around the star z Tauri (Quirrenbach et al.1994).


Struve's proposed Be star model
Figure 1.2: Struve (1931) proposed a Be star model envelope, showing how the observed line profiles depended on viewing angle. The underlying star is heavily shaded, whilst the more diffuse envelope is lightly shaded. (A) corresponds to Figure 1.3 left panel, (B) centre panel and (C), a shell star, the right panel.

A sub-class of the Be stars are shell stars, which are interpreted as Be stars oriented so that the circumstellar disc is viewed edge-on see Figure 1.2. Shell stars have spectra that show Balmer emission with sharp absorption cores, narrow absorption lines of ionized metals, and broad HeI absorption, see Figure 1.3.


H Alpha Line Profiles of 3 B stars
Figure 1.3: Ha Line profiles of three Be stars (see Figure 1.2); left f And (Be 5III, v sini = 113 ± 21kms-1 ), middle HD 173371 (Be 7V, v sini = 271 ± 54kms-1 ), and right HD 179343 (Be 7-8shell, v sini = 224 ± 14kms-1 ). Data extracted from the representative sample (Steele and Negueruela2002).

Observations of Be star discs have shown that they vanish and re-appear in a apparently random fashion with time-scales of 100s of days (e.g.,  Hanuschk et al.1993Hirata1995). Such observations are interpreted as an increase in the density of the circumstellar material, for line generation, and a rarefaction for line dispersal Dachs (1987). A large rarefaction can cause a Be star spectrum to resemble a normal B-star spectrum, before once again entering a Be phase. Figure 1.4 shows observations of Pleione over ~30 year period, the disappearance and reappearance of a Be star disc can be easily seen. To be noted is the sudden fading of the disc whilst the re-growth is evident over a much longer time scale. Be stars may go through periods of disc, no-disc and disc reappearance.


Overall variation of Pleione over 30 years
Figure 1.4: Overall variation of Pleione over ~30 years (Hirata1995).

In Figure 1.4 the closed circular symbols indicate a period when the Be star went through a period of being a shell star. Shell stars are accepted to be Be stars viewed edge on. The reason why a Be star should fluctuate between being a shell star and not is unclear. One possible explanation is that the flaring factor of the disc is not a constant. Therefore if a Be star is viewed from a low-latitude a flared disc would be interpreted as a shell star. Should the flaring decrease then the star would be viewed as a “normal” Be star. The method which would give rise to any change of flaring in the disc is not clear.