Precise radial velocities of giant stars VIII. Testing for the presence of planets with CRIRES Infrared Radial Velocities

Abstract

We have been monitoring 373 very bright (V < 6 mag) G and K giants with high precision optical Doppler spectroscopy for more than a decade at Lick Observatory. Our goal was to discover planetary companions around those stars and to better understand planet formation and evolution around intermediate-mass stars. However, in principle, long-term, g-mode nonradial stellar pulsations or rotating stellar features, such as spots, could effectively mimic a planetary signal in the radial velocity data. Our goal is to compare optical and infrared radial velocities for those stars with periodic radial velocity patterns and to test for consistency of their fitted radial velocity semiamplitudes. Thereby, we distinguish processes intrinsic to the star from orbiting companions as reason for the radial velocity periodicity observed in the optical. Stellar spectra with high spectral resolution have been taken in the H-band with the CRIRES near-infrared spectrograph at ESO's VLT for 20 stars of our Lick survey. Radial velocities are derived using many deep and stable telluric CO2 lines for precise wavelength calibration. We find that the optical and near-infrared radial velocities of the giant stars in our sample are consistent. We present detailed results for eight stars in our sample previously reported to have planets or brown dwarf companions. All eight stars passed the infrared test. We conclude that the planet hypothesis provides the best explanation for the periodic radial velocity patterns observed for these giant stars.