An Exploration of Systematic Errors in Transiting Planets and Their Host Stars

Alison Duck, B. Scott Gaudi, Jason D. Eastman, Joseph E. Rodriguez

Published 2022-09-19Version 1

Transiting planet systems offer the best opportunity to measure the masses and radii of a large sample of planets and their host stars. However, relative photometry and radial velocity measurements alone only constrain the density of the host star. Thus, there is a one-parameter degeneracy in the mass and radius of the host star, and by extension the planet. Several theoretical, semi-empirical, and nearly empirical methods have been used to break this degeneracy and independently measure the mass and radius of the host star and planets(s). As we approach an era of few percent precisions on some of these properties, it is critical to assess whether these different methods are providing accuracies that are of the same order, or better than, the stated statistical precisions. We investigate the differences in the planet parameter estimates inferred when using the Torres empirical relations, YY isochrones, MIST isochrones, and a nearly-direct empirical measurement of the radius of the host star using its spectral energy distribution, effective temperature, and \textit{Gaia} parallax. We focus our analysis on modelling KELT-15b, a fairly typical hot Jupiter, using each of these methods. We globally model TESS photometry, optical-to-NIR flux densities of the host star, and \textit{Gaia} parallaxes, in conjunction with extant KELT ground-based follow-up photometric and radial velocity measurements. We find systematic differences in several of the inferred parameters of the KELT-15 system when using different methods, including a $\sim 6\%$ ($\sim 2\sigma$) difference in the inferred stellar and planetary radii between the MIST isochrones and SED fitting.