Exploring Systematic Errors in the Inferred Parameters of the Transiting Planet KELT-15b and its Host Star

Abstract

Transiting planet systems offer a unique opportunity to measure the masses and radii of many planets and their host stars. Yet, relative photometry and radial velocity measurements alone only constrain the density of the host star. In remedy, the community uses theoretical and semi-empirical methods to break this one-parameter degeneracy and measure the mass and radius of the host star and its planet(s). We investigate the differences in the inferred system parameters due to modeling a host star with the Torres mass-radius relations, YY evolutionary tracks, MIST evolutionary tracks, and a stellar radius estimate from the spectral energy distribution (SED). We consider the effects of different priors on the stellar effective temperature, limb darkening, and eccentricity of the planet. Using the publicly available software package EXOFASTv2, we globally model TESS photometry and radial velocity observations of KELT-15, which hosts a fairly representative hot Jupiter. In total, we explore the impact of 28 different choices of priors on the inferred parameters of KELT-15b. We find broad agreement in the inferred system parameters across methodologies at the level of ~1.1 sigma between the MIST and SED constraints. This gives some confidence that systematic errors are not ubiquitous in transiting planets systems. We also find a ~2 sigma difference in the stellar radius estimated by the MIST models when we adopt differing literature spectroscopic effective temperature estimates. Similar studies of a large number of systems are needed to definitely assess systematic uncertainties the exoplanet population as a whole.