Bayesian inference of stellar parameters based on 1D stellar models coupled with 3D envelopes. (arXiv:1910.04200v1 [astro-ph.SR])

<a href="http://arxiv.org/find/astro-ph/1/au:+Jorgensen_A/0/1/0/all/0/1">Andreas Christ Sølvsten Jørgensen</a>, <a href="http://arxiv.org/find/astro-ph/1/au:+Angelou_G/0/1/0/all/0/1">George C. Angelou</a>

Stellar models utilising one-dimensional (1D), heuristic theories of

convection fail to adequately describe the energy transport in superadiabatic

layers. The improper modelling leads to well-known discrepancies between

observed and predicted oscillation frequencies for stars with convective

envelopes. Recently, three-dimensional (3D) hydrodynamic simulations of stellar

envelopes have been shown to facilitate a realistic depiction of superadiabatic

convection in 1D stellar models. The resulting structural changes of the

boundary layers have been demonstrated to impact not only the predicted

oscillation spectra but evolution tracks as well. In this paper, we quantify

the consequences that the change in boundary conditions has for stellar

parameter estimates of main-sequence stars. For this purpose, we investigate

two benchmark stars, Alpha Centauri A and B, using Bayesian inference. We show

that the improved treatment of turbulent convection makes the obtained 1D

stellar structures nearly insensitive to the mixing length parameter. By using

3D simulations in 1D stellar models, we hence overcome the degeneracy between

the mixing length parameter and other stellar parameters. By lifting this

degeneracy, the inclusion of 3D simulations has the potential to yield more

robust parameter estimates. In this way, a more realistic depiction of

superadiabatic convection has important implications for any field that relies

on stellar models, including the study of the chemical evolution of the Milky

Way Galaxy and exoplanet research.

Stellar models utilising one-dimensional (1D), heuristic theories of

convection fail to adequately describe the energy transport in superadiabatic

layers. The improper modelling leads to well-known discrepancies between

observed and predicted oscillation frequencies for stars with convective

envelopes. Recently, three-dimensional (3D) hydrodynamic simulations of stellar

envelopes have been shown to facilitate a realistic depiction of superadiabatic

convection in 1D stellar models. The resulting structural changes of the

boundary layers have been demonstrated to impact not only the predicted

oscillation spectra but evolution tracks as well. In this paper, we quantify

the consequences that the change in boundary conditions has for stellar

parameter estimates of main-sequence stars. For this purpose, we investigate

two benchmark stars, Alpha Centauri A and B, using Bayesian inference. We show

that the improved treatment of turbulent convection makes the obtained 1D

stellar structures nearly insensitive to the mixing length parameter. By using

3D simulations in 1D stellar models, we hence overcome the degeneracy between

the mixing length parameter and other stellar parameters. By lifting this

degeneracy, the inclusion of 3D simulations has the potential to yield more

robust parameter estimates. In this way, a more realistic depiction of

superadiabatic convection has important implications for any field that relies

on stellar models, including the study of the chemical evolution of the Milky

Way Galaxy and exoplanet research.

http://arxiv.org/icons/sfx.gif