Astron. Astrophys. 341, 527-538 (1999)

1. Introduction

Solar magnetic activity is believed to originate in a dynamo process, resulting from the coupling between convection and differential rotation. It manifests itself in a wide range of observable phenomena including flares, spots, plages, chromospheric network, prominences, etc.

Solar prominences are structures made of condensed cool matter suspended in the corona by magnetic fields. Prominence-like activity has also been observed in RS CVn binary systems (Hall & Ramsey 1992), although this is not a defining characteristic of the group (Gunn & Doyle 1997; Gunn et al. 1997). Recent observations have suggested that an analogue of solar prominences may be an ubiquituous phenomenon in late-type rapid rotators. Cool clouds of neutral material forced to corotate with the underlying star by its magnetic field have been detected in the K dwarf AB Dor (Cameron and Robinson 1989a,b), several G dwarfs in the Per cluster (Collier-Cameron & Woods 1992), the active K dwarf HD 197890 (Jeffries 1993), the active M dwarf HK Aqr (Byrne et al. 1996, hereafter BER96) and the active M dwarf RE 1816+541 (Eibe 1998), all fast rotators. The characteristic dimensions and heights of these structures are considerably larger than solar prominences. It is well known that the degree of magnetic activity in rapidly rotating late-type stars is higher than in the Sun, due to the higher rotation rates and deeper convective regions.

The distribution of rotational velocity with age in cool dwarfs implies spin-down time-scales for rapidly-rotating stars that are too short to be explained by the classical theory for angular momentum loss. In the classical scenario, rotational braking occurs as a result of the interaction between a stellar wind and the ambient coronal field, as it is believed to occur on the Sun. The rotational velocity is expected to decrease monotonically according to the power law, vsin (Skumanich 1972). Recent studies of stellar rotation in a number of nearby open clusters (Stauffer et al. 1984, 1985; Stauffer & Soderblom 1991, and references therein) demonstrates that a more rapid braking mechanism must be introduced in order to justify the observed spin-down time-scales for G-type rapid rotators. In addition, such a mechanism must be dependent on mass in order to explain the longer time-scales that are observed for lower mass stars.

The presence of co-rotating prominences in rapid rotators could account for an additional loss of angular momentum as suggested by Cameron & Robinson (1989a,b). Therefore, it is important to determine how common this phenomenon is, to estimate the amount of mass involved and to derive cloud heights above the stellar surface in order to confirm the significance of its associated angular momentum.

Prominence clouds can be detected spectroscopically because they resonantly scatter the underlying chromospheric radiation out of the line of sight, producing transient absorption features in the rotationally broadened chromospheric H line profile. Since the clouds are forced to corotate with the star by their containing magnetic field, these transients move rapidly across the line profile with monotonically increasing radial velocities resulting in a systematic variability of the H emission. This effect repeats each time a cloud crosses the observer's line of sight in front of the stellar disk.

The star BD + has been selected as a possible candidate for prominence activity. Previous observations of this star (Jeffries et al. 1994, hereafter J94) showed it to be a single K5-K7 dwarf with a projected rotational velocity vsin69 km s^-1. The most probable period as determined from photometric data was found to be 10.170.10hr, which, in combination with an assumed main sequence radius for a star of its spectral type, yields an axial inclination of . Flaring and plage activity were detected but no convincing evidence of prominence-like structures was found. A possible geometric explanation was proposed at the time, related to the fact that co-rotating cool prominences may be impossible to detect in projection on the stellar disk in this particular case because of the low inclination of this star, i.e. that clouds occur only close to the equatorial plane. This view is further reinforced by recent modeling work of van den Oord et al. (1997).

In the following section we describe new observations of this star. Details of the analysis of the data, together with its results are contained in Sect. 3. In Sect. 4 we discuss possible interpretations for the various sources of chromospheric emission variability that have been identified in this system. Finally, the conclusions are summarized in Sect. 5.

© European Southern Observatory (ESO) 1999

Online publication: December 4, 1998
helpdesk.link@springer.de