Deconfusing Detections in Directly Imaged Multiplanet Systems

High-contrast images from future space-based telescopes may contain several planets from multiplanet systems and potentially a few planet-like speckles. When taken several months apart, the short-period planets and speckles will appear to move significantly, to the point that it might not be clear which point source (detection) in the image belongs to which object. In this work, we develop a tool, the deconfuser, to test quickly all the plausible partitions of detections by planets based on orbital mechanics. We then apply the deconfuser to a large set of simulated observations to estimate "confusion" rates, i.e., how often there are multiple distinct orbit combinations that describe the data well. We find that in the absence of missed and false detections, four observations are sufficient to avoid confusion, except for systems with high inclinations (above 75 degrees). In future work, the deconfuser will be integrated into mission simulation tools, such as EXOSIMS, to assess the risk of confusion in missions such as the IR/O/UV large telescope recommended by the Astro2020 decadal survey.