Astron. Astrophys. 364, 665-673 (2000)

3. Observations and data processing

3.1. Radial-velocity observations

We measure radial-velocities for stars in our sample with the ELODIE spectrograph (Baranne et al. 1996) on the 1.93 m telescope of the Observatoire de Haute Provence (France). This fixed configuration dual-fiber-fed echelle spectrograph covers in a single exposure the 390-680 nm spectral range, at an average resolving power of 42000. An elaborate on-line processing is integrated with the spectrograph control software, and automatically produces optimally extracted and wavelength calibrated spectra, with algorithms described in Baranne et al. (1996). All stars in this programme are observed with a Thorium lamp illuminating the monitoring fiber, as needed for the best () radial-velocity precision. The present paper uses data obtained between September 1995 and April 2000.

A few measurements were also obtained with the CORALIE spectrograph on the recently commissioned 1.2-m Euler telescope at La Silla Observatory (Chile). CORALIE is an improved copy of ELODIE and has very similar characteristics, with the exception of a substantially improved intrinsic stability and a somewhat higher spectral resolution ().

These spectra are analysed for velocity by numerical cross-correlation with a one-bit (i.e. 0/1) template. This processing is standard for ELODIE spectra (Queloz 1995a, 1995b). The correlation mask used here was derived by Delfosse et al. (1999c) from a high S/N spectrum of Gl 699 (Barnard's stars, M4V). As discussed in Sect. 4, we determine the orbital parameters of double-lined systems through a direct least square adjustment to the correlation profiles. We recommend that reanalyses of those data similarly use those profiles (available upon request to the authors). For easier reference we nonetheless provide in Tables 4 to 10 (only available in the electronic version of this paper) radial velocities for all stars, obtained from adjustment of Gaussian functions to the correlation profiles. The measurement accuracies for the sources discussed here range between 10 and 100 (depending on apparent magnitude and spectral type), except for the two fastest rotators, GJ 2069 A and YY Gem.

3.2. Adaptive optics imaging

Adaptive optics observations are obtained at the 3.6-meter Canada-France-Hawaii Telescope (CFHT) using PUE'O, the CFHT Adaptive Optics Bonnette (Arsenault et al. 1994, Rigaut et al. 1998) and two different infrared cameras (Nadeau et al. 1994, Doyon et al. 1998). Delfosse et al. (1999c) provide a detailed description of the observing procedure, which we only summarize here.

The program stars are observed in a 4 or 5 positions mosaic pattern, that allows to both determine the sky background from the on-source frames and fully compensate the cosmetic defects of the detector. The science targets are used to sense and correct the incoming wavefront. All of them are bright enough (R  14) to ensure diffraction-limited images in the H and K bands under standard Mauna Kea atmospheric conditions (i.e. for seeing up to 1"). The corrected point spread function obtained from the AO system is synthesized from simultaneous records of the wavefront sensor measurements and deformable mirror commands, as described by Véran et al. (1997). For pre-1997 observations this ancillary information was not available from the acquisition system, and the point spread function is then instead estimated from observations of a reference single star of similar R-band magnitude. Astrometric calibration fields such as the central region of the Trapezium Cluster in the Orion Nebula (McCaughrean & Stauffer 1994), were observed to accurately determine the actual detector plate scale and orientation on the sky.

In good seeing conditions the binaries are observed through J (1.2 µm), H (1.65 µm) and K(2.23 µm) filters, or through corresponding narrow-band filters (usually [Fe⁺] (1.65 µm) and Br (2.166 µm)) for sources which would otherwise saturate the detectors in the minimum available integration time. For worse seeing we restrict observations to the K band, to maintain an acceptable corrected image quality.

We use a deconvolution algorithm (Véran et al. 1999) based on the Levenberg-Marquardt minimisation method and coded within IDL to determine the separation, position angle and magnitude difference between the two stars. With approximate initial values of the positions of the two components along with the PSF reference image, the fitting procedures outputs the flux and pixel coordinates of both stars. The astrometric calibrations then yields the desired angular separations. Tables 11 to 15 (only available electronically) list the individual measurements.

Additional angular separations could be obtained from the litterature for some binaries. They are also listed in Tables 11 to 15, and discussed in Sect. 4.2 for each relevant system

3.3. Parallaxes

As discussed in Sect. 4, the orbital adjustment can make use of the trigonometric parallax of a multiple system, which is handled as an additional observational constraint on the ratio of its physical and angular dimensions. We have obtained this information (Table 1) from the Yale General Catalog of trigonometric Parallaxes (Van Altena et al., 1995) and the HIPPARCOS catalog (ESA 1997), with some individual entries from Probst (1977) and Soderhjelm (1999).

Table 1. Trigonometric parallaxes used for the orbital adjustment. All values are in mas, the references are Van Altena et al. (1995, Yale); ESA (1997, HIPPARCOS catalog); Probst (1977) and Soderhjelm (1999). The HIPPARCOS value listed for Gl 644 corresponds to its common proper motion companion Gl 643, as the measurement for Gl 644 itself is affected by its unaccounted orbital motion.

© European Southern Observatory (ESO) 2000

Online publication: January 29, 2001
helpdesk.link@springer.de