Calibration of the Gaia Radial Velocity Spectrometer from ground-based observations of candidate standard stars

The Radial Velocity Spectrometer (RVS) on board of Gaia will perform a large spectroscopic survey to determine the radial velocities of some 1.5x10^8 stars. We present the status of ground-based observations of a sample of 1420 candidate standard stars designed to calibrate the RVS. Each candidate star has to be observed several times before Gaia launch (and at least once during the mission) to ensure that its radial velocity remains stable during the whole mission. Observations are performed with the high-resolution spectrographs SOPHIE, NARVAL and CORALIE, completed with archival data of the ELODIE and HARPS instruments. The analysis shows that about 7% of the current catalogue exhibits variations larger than the adopted threshold of 300 m/s. Consequently, those stars should be rejected as reference targets, due to the expected accuracy of the Gaia RVS. Emphasis is also put here on our observations of bright asteroids to calibrate the ground-based velocities by a direct comparison with celestial mechanics. It is shown that the radial velocity zero points of SOPHIE, NARVAL and CORALIE are consistent with each other, within the uncertainties. Despite some scatter, their temporal variations remain small with respect to our adopted stability criterion.

The catalog of radial velocity standard stars for the Gaia RVS: status and progress of the observations

A new full-sky catalog of Radial Velocity standard stars is being built for the determination of the Radial Velocity Zero Point of the RVS on board of Gaia. After a careful selection of 1420 candidates matching well defined criteria, we are now observing all of them to verify that they are stable enough over several years to be qualified as reference stars. We present the status of this long-term observing programme on three spectrographs : SOPHIE, NARVAL and CORALIE, complemented by the ELODIE and HARPS archives. Because each instrument has its own zero-point, we observe intensively IAU RV standards and asteroids to homogenize the radial velocity measurements. We can already estimate that ~8% of the candidates have to be rejected because of variations larger than the requested level of 300 m/s.