Confronting uncertainties in stellar physics

Abstract

We assess the systematic uncertainties in stellar evolutionary calculations for low- to intermediate-mass, main-sequence stars. We compare published stellar tracks from several different evolution codes with our own tracks computed using the stellar codes stars and mesa. In particular, we focus on tracks of 1 and 3 M⊙ at solar metallicity. We find that the spread in the available 1 M⊙ tracks (computed before the recent solar composition revision) can be covered by tracks between 0.97−1.01 M⊙ computed with the stars code. We assess some possible causes of the origin of this uncertainty, including how the choice of input physics and the solar constraints used to perform the solar calibration affect the tracks. We find that for a 1 M⊙ track, uncertainties of around 10% in the initial hydrogen abundance and initial metallicity produce around a 2% error in mass. For the 3 M⊙ tracks, there is very little difference between the tracks from the various different stellar codes. The main difference comes in the extent of the main sequence, which we believe results from the different choices of the implementation of convective overshooting in the core. Uncertainties in the initial abundances lead to a 1−2% error in the mass determination. These uncertainties cover only part of the total error budget, which should also include uncertainties in the input physics (e.g., reaction rates, opacities, convective models) and any missing physics (e.g., radiative levitation, rotation, magnetic fields). Uncertainties in stellar surface properties such as luminosity and effective temperature will further reduce the accuracy of any potential mass determinations.