http://arxiv.org/abs/2006.06528
In this work we quantify the effect of an unresolved companion star on the derived stellar parameters of the primary star if a blended spectrum is fit assuming the star is single. Fitting tools that determine stellar parameters from spectra typically fit for a single star, but we know that up to half of all exoplanet host stars may have one or more companion stars. We use high-resolution spectra of planet host stars in the Kepler field from the California-Kepler Survey to create simulated binaries; we select 8 stellar pairs and vary the contribution of the secondary star, then determine stellar parameters with SpecMatch-Emp and compare them to the parameters derived for the primary star alone. We find that in most cases the effective temperature, surface gravity, metallicity, and stellar radius derived from the composite spectrum are within 2-3 $\sigma$ of the values determined from the unblended spectrum, but the deviations depend on the properties of the two stars. Relatively bright companion stars that are similar to the primary star have the largest effect on the derived parameters; in these cases the stellar radii can be overestimated by up to 60%. We find that metallicities are generally underestimated, with values up to 8 times smaller than the typical uncertainty in [Fe/H]. Our study shows that follow-up observations are necessary to detect or set limits on stellar companions of planetary host stars so that stellar (and planet) parameters are as accurate as possible.
E. Furlan and S. Howell
Fri, 12 Jun 20
22/69
Comments: Accepted by ApJ; 20 pages
You must be logged in to post a comment.