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.

中文翻译：

---

来自星震表面校正的 HR 图中的差分建模系统学

演化恒星模型近地表层的局部建模误差导致其正常振荡模式的频率与具有匹配内部结构的实际恒星的频率不同。这些频率差称为星震表面项。通过对单个模式频率的详细约束估计的全球恒星属性先前已被证明相对于该表面项的不同参数化是稳健的。也有人认为这可能适用于更广泛的非参数处理。我们研究了在不同表面项处理方面推断的恒星性质的系统差异，无论是对于统计上大量的主序星样本，还是对于红巨星样本，以前没有对其进行过此类表征。对于主序星，我们证明，虽然质量和半径以及年龄确实对表面项的选择具有鲁棒性，但推断的初始氦丰度 $Y_0$ 对表面校正的选择很敏感。这意味着从详细的星震学返回的氦丰度估计值取决于方法论。另一方面，对于我们的红巨星样本，非参数表面校正返回的推断恒星性质与参数性质截然不同。这些差异的性质表明，这种非参数方法应该更适合演化恒星。这应该在更大的样本上进行验证。推断的初始氦丰度 $Y_0$ 对表面校正的选择很敏感。这意味着从详细的星震学返回的氦丰度估计值取决于方法论。另一方面，对于我们的红巨星样本，非参数表面校正返回的推断恒星性质与参数性质截然不同。这些差异的性质表明，这种非参数方法应该更适合演化恒星。这应该在更大的样本上进行验证。推断的初始氦丰度 $Y_0$ 对表面校正的选择很敏感。这意味着从详细的星震学返回的氦丰度估计值取决于方法论。另一方面，对于我们的红巨星样本，非参数表面校正返回的推断恒星性质与参数性质截然不同。这些差异的性质表明，这种非参数方法应该更适合演化恒星。这应该在更大的样本上进行验证。非参数表面改正返回的推断恒星性质与参数性质大不相同。这些差异的性质表明，这种非参数方法应该更适合演化恒星。这应该在更大的样本上进行验证。非参数表面改正返回的推断恒星性质与参数性质大不相同。这些差异的性质表明，这种非参数方法应该更适合演化恒星。这应该在更大的样本上进行验证。