Accounting for Convective Blue-Shifts in the Determination of Absolute Stellar Radial Velocities

For late-type non-active stars, gravitational redshifts and convective blueshifts are the main source of biases in the determination of radial velocities. If ignored, these effects can introduce systematic errors of the order of ~ 0.5 km/s. We demonstrate that three-dimensional hydrodynamical simulations of solar surface convection can be used to predict the convective blue-shifts of weak spectral lines in solar-like stars to ~ 0.070 km/s. Using accurate trigonometric parallaxes and stellar evolution models, the gravitational redshifts can be constrained with a similar uncertainty, leading to absolute radial velocities accurate to better than ~ 0.1 km/s.

Center-to-Limb Variation of Solar 3-D Hydrodynamical Simulations

We examine closely the solar Center-to-Limb variation of continua and lines and compare observations with predictions from both a 3-D hydrodynamic simulation of the solar surface (provided by M. Asplund and collaborators) and 1-D model atmospheres. Intensities from the 3-D time series are derived by means of the new synthesis code ASSET, which overcomes limitations of previously available codes by including a consistent treatment of scattering and allowing for arbitrarily complex line and continuum opacities. In the continuum, we find very similar discrepancies between synthesis and observation for both types of model atmospheres. This is in contrast to previous studies that used a ``horizontally and time averaged representation of the 3-D model and found a significantly larger disagreement with observations. The presence of temperature and velocity fields in the 3-D simulation provides a significant advantage when it comes to reproduce solar spectral line shapes. Nonetheless, a comparison of observed and synthetic equivalent widths reveals that the 3-D model also predicts more uniform abundances as a function of position angle on the disk. We conclude that the 3-D simulation provides not only a more realistic description of the gas dynamics, but, despite its simplified treatment of the radiation transport, it also predicts reasonably well the observed Center-to-Limb variation, which is indicative of a thermal structure free from significant systematic errors.