Utilizing Astrometric Orbits to Obtain Coronagraphic Images of Extrasolar Planets

Abstract

ABSTRACTWe present an approach for utilizing astrometric orbit information to improve the yield of planetary images and spectra from a follow-on direct-detection mission. This approach is based on the notion—strictly hypothetical—that if a particular star could be observed continuously, the instrument would in time observe all portions of the habitable zone so that no planet residing therein could be missed. This strategy could not be implemented in any realistic mission scenario. But if an exoplanet’s orbit is known from astrometric observation, then it may be possible to plan and schedule a sequence of imaging observations that is the equivalent of continuous observation. A series of images—optimally spaced in time—could be recorded to examine contiguous segments of the orbit. In time, all segments would be examined, leading to the inevitable detection of the planet. In this article, we show how astrometric orbit information can be used to construct such a sequence. We apply this methodology to seven stars taken from the target lists of proposed astrometric and direct-detection missions. In addition, we construct this sequence for the Sun-Earth system as it would appear from a distance of 10 pc. In constructing these sequences, we have assumed that the imaging instrument has an inner working angle (IWA) of 75 mas and that the planets are visible whenever they are separated from their host stars by ≥IWA and are in quarter-phase or greater. In addition, we have assumed that the planets orbit at a distance of 1 AU scaled to luminosity and that the inclination of the orbit plane is 60°. For the individual stars in this target pool, we find that the number of observations in this sequence ranges from two to seven, representing the maximum number of observations required to find the planet. The probable number of observations ranges from 1.5 to 3.1. These results suggest that a direct-detection mission using astrometric orbits would find all eight exoplanets in this target pool with a probability of unity and that the maximum number of visits required (i.e., the worst case) is 36 visits. The probable number of visits is considerably smaller, about 18. This is a dramatic improvement in efficiency over previous methods proposed for utilizing astrometric orbits. We examine how the implementation of this approach is complicated and limited by operational constraints and how it is impacted by formal errors. We find that it can be fully implemented for internal coronagraph and visual nuller missions, with a success rate approaching 100%. External occulter missions will also benefit, but to a lesser degree.