A major program for the VLTI is to study systematically the rich circumstellar environments of YSOs at a resolution of about 2mas, which corresponds to 20-30 stellar radii (0.3 AU) for the nearest star-forming regions (d = 150pc). The factor of twenty increase in resolution over HST provides access to the phenomena which occur in the inner regions around young stars and should provide important input to the theoretical models. However, even VLTI will not be able to resolve the innermost parts of the accretion disk, where material is presumably funnelled via magnetic fields onto the stellar surface and where other parts of the rotating magnetosphere accelerate and collimate the outflowing matter. Nevertheless, observing just outside these regions should allow meaningful extrapolations.
Very few direct studies of circumstellar disks have been performed so far, because this requires high resolution in the near- and mid-infrared domains. Important parameters yet to be determined include the morphology of circumstellar disks, the temperature distribution, the relative contributions from scattered stellar light and thermal disk emission, the disk chemical composition and the properties of dust grains.
In a few objects, minima in the broadband spectrum have been tentatively attributed to zones cleared by a planet or faint companion (Marsh & Mahoney, 1993), although different interpretations based on material properties also are possible. These gaps lie around 1 AU and would be detectable with the VLTI (Malbet & Berthout, 1995). The determination of visibility curves at 2 and 10 microns should indicate the interesting candidates; imaging will be required to study the phenomenon with its asymmetries due to the presence of the orbiting object. Generally, distribution of dust and gas, and of the spatial distribution of temperature can be measured and will clarify the initial conditions for possible subsequent planet formation.
The question of how YSO jets are accelerated and collimated should also
be addressed with the VLTI. Although the innermost region will not be resolved,
important constraints on models can be derived from observations beyond
about 30 stellar radii. A start has been made with HST and ground-based
telescopes, and studies of jet width as a function of radius show that
at least some YSO jets have full opening angles of greater than 50 degrees
for small distances from the star This behaviour, which is observed for
of at least a few YSO jets, is illustrated in Figure 1 on the basis of
a recent HST/WFC image of the bipolar jets from the HH30-star in the [SII]
6716,6731 lines (Ray and Mundt 1996). The figure shows how drastically
larger the opening angle of this YSO jet must be on small scales (i.e.
within < 0.3" or for < 50 AU from the star). A similar behaviour
has been predicted by the theoretical models of Camenzind (1990) in which
the jets are accelerated and collimated by rotating magnetospheres and
in which one expects large jet opening angles for jet radii much smaller
than the light cylinder, which is expected to have a radius of about 30
to 100 AU for typical rotation periods for T Tauri stars of a few days.
Figure 1: The top part shows a HST/WFC image of the bipolar jets from the HH 30-star in the [SII] 6716,6731 lines (with the continuum contribution subtracted). The HH30-star is not visible due to strong extinction in the circumstellar disk. The star is assumed to be located in the center of the gap between the two jets (i.e. 0.3" to the left of the very end of the brighter jet). In the lower part, the measured FWHM of the bipolar jets as a function of distance is shown. For the visible parts of the jet, average full opening angles of about 6 degrees (left part) and 2 degrees (right part) have been derived. However for the invisible part of the jet, i.e. between the source and the first point of measurement the jet opening angle must be considerably larger (at least 50 degrees) as indicated by the dashed line.
The VLTI can also investigate possible connections between variations of the central star and the formation of new knots in the jet. For a jet speed of 300 km/s, a new knot resulting from an outburst would move outwards and be detectable after a few days, allowing its proper motion to be accurately measured. This is similar to VLBI observations of QSO jets. The need to pursue these observations with high spectral resolution (R > 1000) and within 1-2 days because of the high proper motion of the knots (up to 1 mas/day) probably will require the inclusion of the UTs on the basis of current brightness estimates.
Measuring orbits of very close binaries would be another valuable science program. High angular resolution is needed to produce accurate masses for lunar-occultation and spectroscopic binaries within a reasonable time: an orbit of 100AU takes about 1000 years, but one of 4AU will be completed in 8 years. Fortunately, the incidence of binaries among young stars is high (Reipurth & & Zinnecker, 1993) and is probably close to 100% in some molecular clouds (Ghez et al,1993;Leinert et al,1993). Masses from binary orbits will finally provide urgently needed empirical checks on the evolutionary tracks used in interpreting the observations of young stars.
The key capabilities of VLTI in this field are its resolution, sensitivity and infrared response. There are many bright YSOs (V = 11-15, K = 6-9) which will make ideal targets. Once the first two ATs become available, a good first-light project would be to determine stellar masses from orbital motion and to search for new companions. Measuring visibilities at different wavelengths in the near- and mid-infrared bands will allow us to investigate the temperature and density distributions of the circumstellar material and search for gaps which might indicate the formation of planets. Finally, closure-phase imaging will be used to provide the detailed geometrical information needed to understand the fascinating phenomena mentioned above.