Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 352, 574-586 (1999)

Previous Section Next Section Title Page Table of Contents

1. Introduction

Star formation is widely thought to occur in molecular clouds, which then provide the raw material from which stars can accumulate during the course of the gravitational collapse. While many details of the star formation process remain topics of controversy, the idea that most young stellar objects (YSOs) should be physically associated with molecular clouds remained unchallenged until recently.

Two observational findings in the last few years did, however, cast some doubt on this widely held belief. First came the discovery by the ROSAT satellite of a population of X-ray active pre-main sequence stars - many identified as weak-emission line T Tauri stars (WTTSs) - extending far beyond the boundaries of known star-forming regions (e.g., Neuhäuser et al. 1995a, Neuhäuser et al. 1995b). This result was at first taken as an indication that YSOs could migrate far away from the region of their formation on short time-scales. Propositions to explain this fact included the formation of stars in fast-moving molecular cloudlets (Feigelson 1996) or tidally-induced escapes within multiple systems (Armitage & Clarke 1997). However, the recent finding that many of these pre-main sequence objects are most likely the low-mass counterpart of the Gould Belt OB associations (Guillout et al. 1998) reconciles the ROSAT results with the conventional idea that these stars formed in (now dispersed) molecular clouds.

Another observation which may contradict the hypothesis that YSOs and molecular clouds are associated was recently reported by Favata et al. (1998). On the basis of the Hipparcos satellite astrometric measurements, these authors assert that some classical T Tauri stars (CTTSs) of the Taurus-Auriga star-forming region are apparently much closer than previously thought and may not belong to the Taurus clouds. If this result were proven true beyond reasonable doubt, consequences for the physics of star formation and evolution of solar-type stars would be far-reaching, since arguments for the youth of a given stellar object relate primarily to its location in the vicinity of molecular clouds.

Four arguments are used to infer pre-main sequence status of a solar-type star:

  • association with OB stars (e.g., in the Orion Trapezium region);

  • association with dark or bright nebulosity (e.g., in the Taurus-Auriga region);

  • location above the main-sequence in the Hertzsprung-Russell Diagram;

  • presence in the spectrum of the [FORMULA] Å LiI resonance line (with equivalent width larger than in ZAMS stars of the same spectral type).

The first two criteria are straightforward: OB associations are short-lived, which guarantees the youth of associated low-mass objects, while physical association of a star with a cloud is either seen at the telescope when close reflection and emission nebulae are present - as is the case in the vicinity of many CTTSs - or inferred from kinematic studies (e.g. Jones & Herbig 1979).

The last two criteria above are more indirect. The location in the H-R diagram depends on the assumption made about the respective luminosities of photosphere and circumstellar matter in a given object (Kenyon & Hartmann 1990), as well as on the assumed distance. Lithium is, in theory, destroyed early in the star history, when the temperature at the bottom of the convection zone reaches about [FORMULA] K, but the lithium abundance may be dependent on variables other than age (cf., e.g., Ventura et al. 1998).

If some CTTSs appear to be located nearby, and thus far away from known star-forming region, as suggested by Favata et al. (1998), one must then choose between two equally unsettling possibilities: (a) either they are young stars, and one must explain how they arrived at their present location, or (b) they are field stars that mimic YSOs, and one must understand how this is possible. Needless to say, there are no obvious answers to these questions, and the entire picture of low-mass star formation would have to be fully revised in order to account for these observations.

This paper provides a detailed re-examination of Hipparcos satellite data for TTSs, focusing mainly on distances and binarity/multiplicity properties. Data are discussed in Sect. 2. Sect. 3 then examines the Hipparcos results relevant to YSO and molecular cloud distances, and concludes that Hipparcos distances for young stars are generally consistent with their expected physical association with molecular clouds. The same conclusion was reached independently by Wichmann et al. (1998), but we extend their analysis by providing new astrometric solutions for groups of TTSs in various star-forming regions. We then study the binarity/multiplicity Hipparcos data in Sect. 4. There, we discuss or improve the astrometric data reductions for several YSO systems, using Hipparcos intermediate astrometric data and, when available, taking advantage of recent spectroscopic information.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: December 2, 1999