Astron. Astrophys. 331, 581-595 (1998)

1. Introduction

Surface rotation is a key observational parameter for stellar evolution, as a diagnostic of the mechanisms responsible for stellar angular momentum loss and internal angular momentum transport. The latter are responsible for chemical mixing and they modify the stellar thermal structure (e.g. Martin & Claret 1996), and hence affect the overall evolution. Rotation is also the driving force behind stellar activity (coronae, chromospheres, spots, flares), through dynamo production of magnetic fields.

Pre-main-sequence rotational evolution of low mass stars is generaly presumed to be dominated by magnetic coupling with the parent protostellar disk (Cameron & Campbell 1993, Shu et al. 1994): as long as a star is surrounded by a substantial accretion disk, its equatorial velocity remains approximately constant at  20  (e.g. Bouvier et al. 1993). Once accretion stops, disk braking disappears, and angular momentum conservation then becomes the dominant factor. Stars therefore spin up during their contraction along an Hayashi track, as their moment of inertia decreases. Since the time at which this happens depends on details of their circumstellar environment, stars arrive on the ZAMS with a broad distribution of rotational velocities (e.g. Bouvier et al. 1997). Once on the main sequence the moment of inertia no longer changes significantly. As first suggested by Schatzman (1962), magnetic winds then brake down the stars to low final equatorial velocities, approximately with a Skumanich (1972) law. The time scale for this angular momentum dissipation is mass-dependent: rapid rotators are found at all spectral types (G-M) at the age of the Persei cluster ( 50 Myr) (Prosser 1991), in the Pleiades ( 70 Myr) only later than mid-K (Stauffer & Hartmann, 1987), and in the Hyades ( 600 Myr) only within the M dwarfs (Stauffer et al. 1987; Stauffer et al. 1997). The situation in the field is slightly less clear, since several age groups are represented, but all G dwarfs rotate slowly, and there are good spectroscopic or dynamic arguments to attribute a (very) young age to all K-early M rapid rotators.

Over the years, a large observational effort has established this picture of rotation along low mass stellar tracks, and models incorporating the above general elements (e.g. Bouvier & Forestini 1995, Krishnamurthi et al. 1997) successfully reproduce the rotational velocity distribution of solar mass dwarfs at all ages: pre-main-sequence T Tauri and post-T Tauri star, young main sequence stars in open clusters, and field stars.

Some aspects of the models however remain largely phenomenological, mostly because the complex physics of accretion and stellar dynamos is only partly understood. A number of competing models therefore exist, that differ on, for instance, radiative core/convective envelope decoupling, or parameterisation of the angular momentum loss law. The uncertain physical mechanisms are expected to have a strong mass dependence, and it is therefore useful to extend the observational database to the lower mass M dwarfs (e.g. Krishnamurthi et al. 1997). The [M0V,M6V] range is also interesting because it contains the mass (, M2.5V) below which main sequence stars no longer develop a radiative core. The radiative/convective boundary layer is essential (e.g. Spiegel & Weiss 1980; Spruit & van Ballegooijen 1982) to the operation of the standard shell dynamo which is generally believed to generate the large scale solar magnetic field, and a break in both the activity level and the rotation properties could, at least naively, be expected at this spectral type.

A number of authors have discussed the rotation behaviour of M dwarfs in young open clusters (most recently Stauffer et al. (1997) for the Hyades, Stauffer et al. (1994) for the Pleiades, Patten & Simon (1996) for IC 2391), but there are fewer studies for field M dwarfs. Stauffer & Hartmann (1986) measured v sin i for approximately 200 field M dwarfs with a 10 km.s ^-1 sensitivity. Only 11 have significant broadening, a few of which may actually be unrecognized double-lined spectroscopic binaries. Marcy & Chen (1992) observed 47 field M dwarfs with v sin i sensitivity of 3 km.s ^-1, only five of which have detectable rotation, all with v sin i km.s ^-1. Both surveys have relatively bright limiting magnitudes (V=12 and V=11, respectively) and as a consequence preferentially sample early M dwarfs. More recently, two papers have examined the rotation of very low mass field stars (M6 or later). Martin et al. (1996) have determined rotational periods for a set of very late (M6-M9) dwarfs, and all 6 field stars in their sample have very short rotational periods 8 hours (). Basri & Marcy (1995) have measured v sin i for 5 extreme cool field dwarfs, of which 3 rotate, including the brown dwarf candidate BRI 0021-0214 (M9.5+, v sin i =40 km.s ^-1). Surprisingly, BRI 0021-0214 is a rapid rotator but has very weak (Basri & Marcy 1995) though detectable (Tinney et al. 1997) chromospheric activity, as measured by emission. This suggests a possible change in the rotation/activity relation for the latest M dwarfs, and it is therefore important to examine slightly earlier M dwarfs.

In the course of a radial velocity survey for low mass companions and planets around nearby M dwarfs (Delfosse et al. 1997), we have obtained high resolution optical spectra for a volume-limited sample of 118 K5 to M6.5 dwarfs. Here we analyse the accurate v sin i measurements (or significant upper limits, 2 ) that we derive from those data, using digital correlation techniques. They fill the spectral type gap between the generally earlier slow rotators measured by Marcy & Chen (1992) and the later than M6 very low mass rapid rotators observed by Basri & Marcy (1995) and Martin et al. (1996). This brings important new constraints on angular momentum dissipation in mid-M dwarfs at a few Gyr.

© European Southern Observatory (ESO) 1998

Online publication: February 16, 1998
helpdesk.link@springer.de