Surface-effect corrections for solar-like oscillations using 3D hydrodynamical simulations

TL;DR

This study uses 3D hydrodynamical simulations to physically model surface effects in solar-like stars, improving the correction of systematic frequency errors in stellar oscillation data.

Contribution

It provides a physical basis for surface-effect corrections by modeling outer stellar layers with 3D simulations and constrains the correction parameters with new formulations.

Findings

Surface effects depend on effective temperature and surface gravity.

Lorentzian correction formulation is more robust across frequencies.

Scaling relations link correction parameters to stellar surface properties.

Abstract

The space-borne missions have provided us with a wealth of high-quality observational data that allows for seismic inferences of stellar interiors. This requires the computation of precise and accurate theoretical frequencies, but imperfect modeling of the uppermost stellar layers introduces systematic errors. To overcome this problem, an empirical correction has been introduced by Kjeldsen et al. (2008, ApJ, 683, L175) and is now commonly used for seismic inferences. Nevertheless, we still lack a physical justification allowing for the quantification of the surface-effect corrections. We used a grid of these simulations computed with the CO$^5$BOLD code to model the outer layers of solar-like stars. Upper layers of the corresponding 1D standard models were then replaced by the layers obtained from the horizontally averaged 3D models. The frequency differences between these patched models and the 1D standard models were then calculated using the adiabatic approximation and allowed us to constrain the Kjeldsen et al. power law, as well as a Lorentzian formulation. We find that the surface effects on modal frequencies depend significantly on both the effective temperature and the surface gravity. We further provide the variation in the parameters related to the surface-effect corrections using their power law as well as a Lorentzian formulation. Scaling relations between these parameters and the elevation (related to the Mach number) is also provided. The Lorentzian formulation is shown to be more robust for the whole frequency spectrum, while the power law is not suitable for the frequency shifts in the frequency range above $\nu_{\rm max}$.