http://arxiv.org/abs/1704.02493
(abridged) Gaia should increase significantly our knowledge of exoplanet systems, in terms of both their number and physical properties. We develop Bayesian methods and detection criteria for orbital fitting, and revise the detectability of exoplanets in light of the in-flight properties of Gaia. We generate four sets of simulations of Gaia data for exoplanet, considering both the nominal mission length (5 yr), and an extended mission length (10 yr). We fit our simulations using the Stan software for Bayesian inference. We investigate the detection rate according to four different indicators: the Akaike information criterion (AIC), the Bayesian information criterion (BIC), the widely-applicable information criterion (WAIC), and the Delta chi^2. We calibrate detection thresholds that select essentially no false positives, and can detect 70–75% of planets at high SN. All four detection criteria can detect planets with SN~1–1.5 for the 5 yr mission, and 0.7–1 for the 10 yr mission. This sets the maximum distance to which a planet is detectable to ~160 pc and ~8 pc for a Jovian and Neptunian planet, respectively, assuming an extended 10 yr mission, a 4 au semi-major axis, and a 0.5 M_sun star. All orbital parameters are recovered, and we present a summary of the characteristic size of the error bars of each parameter. The period can be determined with the best accuracy, with a median relative error of the order of +-1.3%. The angular dimension of the semi-major axis of the orbit can be recovered with a median relative error of +-17%. The eccentricity can also be recovered with a median absolute error of +-0.14. Model selection based on information criteria is an effective way to decide whether a planet is detected; with the WAIC being better suited then the AIC or the BIC in some circumstances.
P. Ranalli, D. Hobbs and L. Lindegren
Tue, 11 Apr 17
21/62
Comments: 22 pages, 19 figures. Submitted to A&A