2. Previous related work
Catala & Kunasz (1987) proposed a quantitative model of the wind of AB Aur. This model involves a spherically symmetric wind with a mass loss rate of about yr-1, with velocities reaching up to 300-400 km s-1, and an extended chromosphere at the base of the wind, with temperatures as high as 17,000 K (whereas =10,000 K).
Praderie et al. (1986) and Catala et al. (1986b) report a modulation of the Mg II resonance lines at 2800 Å, which are formed in the wind, with a period of 45 hrs, and of the Ca II K line, formed near the photosphere at the base of the wind, with a period of 32 hrs. These periodic variations were interpreted in terms of rotational modulation by these authors. In their model, a surface magnetic field creates an alternation of fast and slow streams in the wind, thus leading to a periodic modulation of the lines formed in the wind with the star's rotation period. The difference between the period in the Mg II line variations and that of the Ca II K line is difficult to understand in the framework of this model. It was tentatively attributed by these authors to the fact that these lines are not formed at the same distance from the star: the Ca II K line, formed very close to the photosphere, is modulated by the star's rotation, while the Mg II lines, formed much further out in the wind, are modulated by the rotation of the envelope at that distance.
Recent observations with the GHRS onboard the HST by Bouret et al. (1997) have revealed the presence of N V resonance lines near 1240 Å. These lines indicate temperatures above 100,000 K. These authors show that they can be formed within the co-rotating interaction regions which are expected at the interface between fast and slow streams, and which may also be responsible for the observed X-ray flux (Zinnecker & Preibisch, 1994). An interesting alternative to this interpretation would involve a magnetically confined wind as suggested for the Ap star IQ Aur by Babel & Montmerle (1997): the N V resonance lines and the X-ray emission would originate from a post-shock region in the magnetic equatorial plane where magnetically channelled streams from the two hemispheres collide.
The H line of AB Aur was also reported to vary from a type II P Cygni profile to a single-peak emission profile (Beskrovnaya et al. 1991, 1995). The most attractive model to explain this type of behavior is that of Pogodin (1992), involving an equatorial wind. In this model, the wind is confined to equatorial regions by a magnetic field, with a variable opening angle. When the line of sight intercepts the wind region, a P Cygni profile is formed, whereas a single-peak emission is produced when it does not. This model has a lot in common with that of Babel & Montmerle (1997) for IQ Aur.
AB Aur was monitored in the He I 5876 Å line during the MUSICOS 92 campaign. The bad weather experienced during the 1992 campaign prevented us from reaching firm conclusions. However, the data show a spectacular variability of this line (Böhm et al. 1996). Whether the variations are periodic or not could not be firmly concluded on the basis of these data, although they present some indication of a periodicity near 34 hours. A high level of short-term variability is also present in addition to the possible periodic modulation. Finally, some low level variability was also discovered in the photospheric lines of AB Aur during the MUSICOS 92 campaign (Catala et al. 1997). Again, the data were not sufficient to conclude anything on the periodicity of these variations.
Vigneron et al. (1990), then more recently Lignières et al. (1996), have studied the effect of the wind on the internal structure and rotation of Herbig Ae stars. They conclude that the torque exerted by the loss of angular momentum at the star's surface excites 3D turbulence in the sub-photospheric layers. These 3D turbulent motions create a mixing layer, which rotates at a slower rate than the inner regions of the star, and which tends to deepen in a typical time scale of years. Now, because the angular momentum loss is highest in the equatorial regions, this effect is maximum at the equator, so that the equator is expected to rotate more slowly than the poles.
Most of the ideas presented above assume the presence of a surface magnetic field, which is responsible for structuring the wind both in latitude and in longitude. In a first attempt using spectropolarimetric techniques, Catala et al. (1993) failed to detect this field, yielding an upper limit of about 1000 G for its intensity. The equipartition field at AB Aur's photosphere being of the order of 100 G, this negative result still left a good margin for the models presented above.
Following these previous results, the goals of the MUSICOS 96 observations of AB Aur were twofold: (i) monitor simultaneously lines formed in the photosphere and in various regions of the wind, in order to obtain constraints on the structure of the photosphere/wind complex; and (ii) attempt a direct detection of a surface magnetic field by Zeeman spectropolarimetry.
© European Southern Observatory (ESO) 1999
Online publication: April 28, 1999