Determination of distances is a fundamental problem in investigations of objects within our galaxy. Spectroscopic or kinematic distances have a high degree of uncertainty, and that uncertainty is generally propagated through to yield even higher uncertainties in astrophysical properties. The large increase in the number of stellar detections promised by the upgrade in sensitivity, coupled with the factor of 7 improvement in the resolving power of the A+ configuration, raises the possibility of carrying out stellar astrometry on a large number of radio stars with sub-milli-arc-second accuracy. Astrometric accuracy at this level would enable accurate parallax measurements out to distances of a large fraction of a kpc. It would also allow the direct measurement of binary motions for nearby non-thermal radio binaries such as RS CVn and Algol systems, and weak thermal emitters such as Be star binaries. Such measurements can yield reliable, independent information on the nature of the binary components and the origin of the radio emission.
One of the chief problems with using stellar radio positions to tie the optical and radio coordinate frames is the binary nature and highly degree of variability of the bright non-thermal radio sources for which accurate radio astrometry is currently available. Thermal emission from the photospheres and winds of single stars is quasi-steady state and centered on the position of the star. These objects are ideally suited for matching optical and radio astrometry. Combined with the observations of bright stars from the Hipparcos mission, radio astrometric positions for a large number of thermal emitting, single stars will provide an extensive grid of accurate measurements to tie the optical and radio reference frames with milli-arc-second precision.