Gravitational redshifts in main-sequence and giant stars

PASQUINI, L.; MELO, C.; CHAVERO, C.; DRAVINS, D.; LUDWIG, H.-G.; BONIFACIO, P; DE LA REZA, R.

ASTRONOMY AND ASTROPHYSICS

EDP SCIENCES S A

Año: 2011 vol. 526 p. 1 - 15

0004-6361

Context. Precise analyses of stellar radial velocities is able to reveal intrinsic causes of the wavelength shifts of spectral lines (other than Doppler shifts due to radial motion), such as gravitational redshifts and convective blueshifts. Aims: Gravitational redshifts in solar-type main-sequence stars are expected to be some 500 m s-1 greater than those in giants. We search for this difference in redshifts among groups of open-cluster stars that share the same average space motion and thus have the same average Doppler shift. Methods: We observed 144 main-sequence stars and cool giants in the M 67 open cluster using the ESO FEROS spectrograph and obtained radial velocities by means of cross-correlation with a spectral template. Binaries and doubtful members were not analyzed, and average spectra were created for different classes of stars. Results: The M 67 dwarf and giant radial-velocity distributions are each well represented by Gaussian functions, which share the same apparent average radial velocity to within 100 m s-1. In addition, dwarfs in M 67 appear to be dynamically hotter (sigma = 0.90 km s-1) than giants (sigma = 0.68 km s-1). Conclusions: We fail to detect any difference in the gravitational redshifts of giants and MS stars. This is probably because of the differential wavelength shifts produced by the different hydrodynamics of dwarf and giant atmospheres. Radial-velocity differences measured between unblended lines in averaged spectra vary with line-strength: stronger lines are more blueshifted in dwarfs than in giants, apparently removing any effect of the gravitational redshift. Synthetic high-resolution spectra are computed from three dimensional (3D) hydrodynamic model atmospheres for both giants and dwarfs, and synthetic wavelength shifts obtained. In agreement with observations, 3D models predict substantially smaller wavelength-shift differences than expected from gravitational redshifts only. The procedures developed could be used to test 3D models for different classes of stars, but will ultimately require high-fidelity spectra for measurements of wavelength shifts in individual spectral lines.