The Challenges of Coronagraphic Astrometry

Abstract

A coronagraph in conjunction with adaptive optics provides an effective means to image faint companions of nearby stars from the ground. The images from such a system are complex, however, and need to be fully characterized and understood before planets or disks can be detected against the glare from the host star. Using data from the Lyot Project coronagraph, we investigate the difficulties of astrometric measurements in diffraction-limited coronagraphic images and consider the principal problem of determining the precise location of the occulted star. We demonstrate how the image structure varies when the star is decentered from the optical axis and show how even small offsets (0.05?/D or 5 mas) give rise to false sources in the image. We consider methods of determining the star position from centroiding, instrument feedback, and analysis of point-spread function symmetry and conclude that internal metrology is the most effective technique. ? 2006. The American Astronomical Society. All rights reserved.