2015A&A...574A.124D

Many large stellar surveys have been and are still being carried out, providing huge amounts of data, for which stellar physical parameters will be derived. Solar twins and analogues provide a means to test the calibration of these stellar catalogues because the Sun is the best-studied star and provides precise fundamental parameters. Solar twins should be centred on the solar values. This spectroscopic study of solar analogues selected from the Geneva-Copenhagen Survey (GCS) at a resolution of 48000 provides effective temperatures and metallicities for these stars. We test whether our spectroscopic parameters, as well as the previous photometric calibrations, are properly centred on the Sun. In addition, we search for more solar twins in our sample. The methods used in this work are based on literature methods for solar twin searches and on methods we developed in previous work to distinguish the metallicity-temperature degeneracies in the differential comparison of spectra of solar analogues versus a reference solar reflection spectrum. We derive spectroscopic parameters for 148 solar analogues (about 70 are new entries to the literature) and verify with a-posteriori differential tests that our values are well-centred on the solar values. We use our dataset to assess the two alternative calibrations of the GCS parameters; our methods favour the latest revision. We show that the choice of spectral line list or the choice of asteroid or time of observation does not affect the results. We also identify seven solar twins in our sample, three of which are published here for the first time. Our methods provide an independent means to differentially test the calibration of stellar catalogues around the values of a well-known benchmark star, which makes our work interesting for calibration tests of upcoming Galactic surveys.