Scheduling Direct Imaging Observations Based on Radial Velocity Orbital Fits: Best Practices for Translating Orbits and Failure Modes

Abstract

Abstract Future direct imaging mission concepts are planning to observe planets discovered via Doppler spectroscopy, with observation scheduling decisions based on radial velocity (RV) fit orbits to increase the mission’s odds of making a detection. However, some orbital parameters, such as inclination and planet radius, that are necessary to determine when a planet is detectable via imaging cannot be resolved with only RV data. This work studies how to best use a RV fit to determine when a planet will be detectable with direct imaging and quantify the impacts of RV-measurement precision on direct imaging observations. Beginning by defining a “true” planet, we simulate RV observations, fit the RV observations via Bayesian orbit fitting, construct many Keplerian orbits based on the RV fit and priors on the unknown orbital parameters, and propagate the constructed orbits and the true orbit for 20 yr to determine when the constructed orbits deviate significantly. We identify four methods of constructing orbits, ultimately finding that creating and sampling high-likelihood orbital parameters from a multivariate Gaussian produces the best results for high RV error and that creating and sampling from a simple normal distribution for each orbital parameter independently performs the best for low RV error. Further, we establish two modes of failure: intermittent failure, which captures when the constructed orbits incorrectly indicate that a planet is detectable; and dispersion failure, when the constructed orbits have dispersed so much that they no longer provide useful information.