Differential Modelling Systematics across the HR Diagram from Asteroseismic Surface Corrections

TL;DR

This study investigates how different surface correction methods affect the inferred stellar properties from asteroseismology, revealing robustness in some parameters for main-sequence stars but significant differences for red giants.

Contribution

It provides a comparative analysis of parametric and nonparametric surface correction methods across different stellar evolutionary stages, highlighting their impact on inferred stellar properties.

Findings

Masses and radii are robust to surface correction choice in main-sequence stars.

Initial helium abundance is sensitive to the surface correction method.

Nonparametric corrections yield significantly different results for red giants.

Abstract

Localised modelling error in the near-surface layers of evolutionary stellar models causes the frequencies of their normal modes of oscillation to differ from those of actual stars with matching interior structures. These frequency differences are referred to as the asteroseismic surface term. Global stellar properties estimated via detailed constraints on individual mode frequencies have previously been shown to be robust with respect to different parameterisations of this surface term. It has also been suggested that this may be true of a broader class of nonparametric treatments. We examine systematic differences in inferred stellar properties with respect to different surface-term treatments, both for a statistically large sample of main-sequence stars, as well as for a sample of red giants, for which no such characterisation has previously been done. For main-sequence stars, we demonstrate that while masses and radii, and hence ages, are indeed robust to the choice of surface term, the inferred initial helium abundance $Y_0$ is sensitive to the choice of surface correction. This implies that helium-abundance estimates returned from detailed asteroseismology are methodology-dependent. On the other hand, for our red giant sample, nonparametric surface corrections return dramatically different inferred stellar properties than parametric ones. The nature of these differences suggests that such nonparametric methods should be preferred for evolved stars; this should be verified on a larger sample.