Research output: Book/anthology/dissertation/report › Ph.D. thesis

- Magnus Johan Aarslev

Asteroseismology is the study of stars through their pulsations, the frequencies of which are determined by the structure and dynamics inside the star. By constructing stellar models whose global and seismic properties match observations, we obtain information about otherwise unobservable interiors of stars. For stars like the sun, energy transport in the outer layers occurs mainly through turbulent convection. Here, pressure mode oscillations are essentially propagating sound waves, whose properties can be altered by interaction with the turbulent motion of the gas. This has always been a problem for asteroseismology, because of the challenges inherent in modelling turbulent convection in 1D stellar models. As a result of oversimplifying the physics near the surface, theoretical calculations systematically overestimate the oscillation frequencies. This has become known as the asteroseismic surface effect. Due to lacking better options, this frequency difference is typically corrected for with ad-hoc formulae.

The topic of this thesis is the improvement of 1D stellar convection models and the effects this has on asteroseismic properties. The source of improvements is 3D simulations of radiation-coupled hydrodynamics in stellar atmospheres. Such models offer a realistic representation of stellar surface convection and perform well when confronted with observations.

Convection simulations are used to improve stellar convection in two ways. First, I present results using temporal and horizontal averages of 3D atmospheres to replace the outer layers of stellar models. The additional turbulent pressure and asymmetrical opacity effects in the atmosphere model, compared to convection in stellar evolution models, serve to expand the atmosphere. The enlarged acoustic cavity lowers the pulsation frequencies bringing them closer to observations. We call this the structural part of the surface effect. I have explored this across the HR-diagram and found that the relative surface effect actually increases with evolution and effective temperature. We test the applicability of different surface effect corrections and find that an inertia-scaled, two-term correction best describes the surface term. In another study, I focused on the impact of correction formulae on the derived stellar parameters, when fitting models to high-quality observations.

The second use of convection simulations is for calibrating parameters of 1D convection models. I present stellar evolution models with mixing-length parameters calibrated to 3D atmosphere models and \ttau relations calculated directly from these. The result is models that are slightly hotter and more compact, which affects the internal structure, the pace of evolution and the frequencies. However, the effects are barely prominent enough to be distinguishable with today's observational precision. But it does provide means of determining the mixing-length and enables consistent patching.

The previously mentioned investigations are based on adiabatic frequency calculations, which neglect energy exchange between convection and pulsations, i.e. the modal part of the surface effect. Studying excitation and damping mechanisms requires a non-adiabatic treatment. A major part of my research has been modelling damping rates of red giant stars observed by {\Kp}. The basis for the non-adiabatic calculations is a nonlocal, time-dependent convection model already proven successful for solar-like, main-sequence stars. The model contains free parameters, some of which are determined from calibration to 3D convection simulations. The results constitute the first proper matches between theoretical frequency-dependent damping rates and observed linewidths and pave the way for a better understanding of the physics and dynamics of giant star envelopes.

The topic of this thesis is the improvement of 1D stellar convection models and the effects this has on asteroseismic properties. The source of improvements is 3D simulations of radiation-coupled hydrodynamics in stellar atmospheres. Such models offer a realistic representation of stellar surface convection and perform well when confronted with observations.

Convection simulations are used to improve stellar convection in two ways. First, I present results using temporal and horizontal averages of 3D atmospheres to replace the outer layers of stellar models. The additional turbulent pressure and asymmetrical opacity effects in the atmosphere model, compared to convection in stellar evolution models, serve to expand the atmosphere. The enlarged acoustic cavity lowers the pulsation frequencies bringing them closer to observations. We call this the structural part of the surface effect. I have explored this across the HR-diagram and found that the relative surface effect actually increases with evolution and effective temperature. We test the applicability of different surface effect corrections and find that an inertia-scaled, two-term correction best describes the surface term. In another study, I focused on the impact of correction formulae on the derived stellar parameters, when fitting models to high-quality observations.

The second use of convection simulations is for calibrating parameters of 1D convection models. I present stellar evolution models with mixing-length parameters calibrated to 3D atmosphere models and \ttau relations calculated directly from these. The result is models that are slightly hotter and more compact, which affects the internal structure, the pace of evolution and the frequencies. However, the effects are barely prominent enough to be distinguishable with today's observational precision. But it does provide means of determining the mixing-length and enables consistent patching.

The previously mentioned investigations are based on adiabatic frequency calculations, which neglect energy exchange between convection and pulsations, i.e. the modal part of the surface effect. Studying excitation and damping mechanisms requires a non-adiabatic treatment. A major part of my research has been modelling damping rates of red giant stars observed by {\Kp}. The basis for the non-adiabatic calculations is a nonlocal, time-dependent convection model already proven successful for solar-like, main-sequence stars. The model contains free parameters, some of which are determined from calibration to 3D convection simulations. The results constitute the first proper matches between theoretical frequency-dependent damping rates and observed linewidths and pave the way for a better understanding of the physics and dynamics of giant star envelopes.

Original language | English |
---|

Number of pages | 129 |
---|---|

State | Published - 14 Dec 2017 |

- Astrophysics - Solar and Stellar Astrophysics, ASTEROSEISMOLOGY, convection, stars: atmospheres, stars: evolution, hydrodynamics,Sun:helioseismology,interior,oscilla

See relations at Aarhus University Citationformats

ID: 119497556