Presentation #408.05 in the session Exoplanet Direct Imaging — iPoster Session.
Space-based telescopes, such as the large IR/O/UV strategic mission recommended in Astro2020, will be able to directly image multiple short-period planets simultaneously. The apparent separations of such planets may vary significantly between observations, especially in high inclination systems. This might lead to confusion: multiple substantially different orbits can fit the observations well. To assess the risk of confusion we have developed an efficient algorithm (the “deconfuser”) for orbit fitting and assigning detections to specific planets. The deconfuser accepts unlabeled astrometric measurements that do not specify which detected object belongs to which planet and potentially even include false detection. It guarantees that all orbits that fit astrometry within specified tolerance are found. We apply the deconfuser on a few thousand Monte-Carlo sampled systems between 0.5 and 6 AU to estimate rates of confusion using astrometry-only data (with tolerances between 0.025 and 0.2 AU). We present several modes of confusion that are prevalent. For example, in systems with three or fewer detections per planet, confusion rates (not being able to decide which detection corresponds to which planet based on astrometry alone) range from few percent to tens of percent. Edge-on systems have significant confusion rates even at four detections per planet.