Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 345, 884-904 (1999)

Previous Section Next Section Title Page Table of Contents

3. Observations and data reduction

In this section, we describe the instruments used for the MUSICOS 96 campaign, as well as the reduction procedures followed for the AB Aur data. Table 1 shows the participating sites and gives a summary of the instrumentation used.


Table 1. Instrument characteristics

All the instruments used during this campaign were cross-dispersed echelle spectrographs. The data obtained at OHP, Hawaii, and La Palma cover a very wide wavelength domain, giving access to many photospheric lines, mainly in the blue, and to several lines formed in the wind and chromosphere, such as He I 5876 Å, H[FORMULA], and Fe II 5018 Å. The spectrographs in use at Xinglong and McDonald cover a narrower spectral range, but sufficiently wide to contain all wind and chromosphere lines of interest.

The MUSICOS spectropolarimeter (Donati et al. 1998) was transported to Hawaii to be used on the 3.6m CFHT. This particular instrumental setup is designed for the study of stellar magnetic fields through the measurement of linear (Stokes Q and U) and circular polarisation (Stokes V) Zeeman signatures in line profiles.

In addition to the main sites and instruments cited above, some data were also collected at Ritter Observatory, with the 1m telescope, equipped with an echelle spectrograph. Unfortunately, these few spectra of AB Aur were obtained in poor weather conditions, resulting in low signal-to-noise ratios, and could not be used in the analysis below.

The weather was not particularly good during the MUSICOS 96 campaign, especially in Hawaii. However, due to the redundancy of the longitude coverage achieved for this campaign, the AB Aur observations cover about 200 hours, with a duty cycle close to 80% for the first 100 hours, and approximately 40% for the remaining 100 hours.

Table 2 presents the log of the observations. In this table, `ohp' stands for Observatoire de Haute Provence, `mdo' for McDonald Observatory, `cfh' for Canada-France-Hawaii telescope, `xlo' for Xinglong, and `int' for Isaak Newton telescope.


Table 2. Log of the MUSICOS 96 observations: the dates refer to Nov. 1996, the UT times are in decimal hours, and the exposure times are in seconds


Table 2. Log of the MUSICOS 96 observations: the dates refer to Nov. 1996, the UT times are in decimal hours, and the exposure times are in seconds

The data obtained from Xinglong, McDonald, and the INT were reduced with the "Esprit" reduction software developed by one of us (Donati et al. 1997). In this method, the position of the echelle orders is automatically detected. The images are corrected for pixel-to-pixel inhomogeneities and blaze function by dividing the images containing the stellar spectra by images of flat-field spectra, after flattening the flat-field frame in the direction perpendicular to the spectrograph dispersion. The signal along the orders of the spectrograms is then extracted using the optimal extraction algorithm (Horne 1986, Marsh 1989). The Th/Ar spectra are extracted by simply summing the data about the central location of each order perpendicularly to the order. The wavelength calibration procedure consists basically in a 2D polynomial fit of thorium and argon lines identified in the spectrum (i.e. both in the direction of the grating dispersion and in the direction of the prism cross-dispersion). All details on this reduction procedure can be found in Donati et al. (1997).

The data obtained at OHP were reduced on-site, using the automatic INTER-TACOS procedure (Baranne et al. 1995).

Finally, the spectra obtained at CFHT with the MUSICOS polarimeter were reduced following a dedicated procedure for extracting Stokes V & I parameters (Donati et al. 1997).

The wavelength scales of the stellar spectra were subsequently transformed to a frame linked to the interstellar Na I D lines present in all of our spectra, by simply setting the measured wavelengths of these lines to their laboratory wavelengths. This procedure allows us to compensate for potential systematic wavelength calibration differences between the sites involved in the campaign. We measured in our spectra an heliocentric velocity of +16 km s-1 for the Na I D interstellar lines, while the star systemic heliocentric velocity is +21 km s-1 (Finkenzeller & Jankovics 1984). The star has a small (+5 km s-1) and constant velocity in the reference frame of the Na I D interstellar lines (Finkenzeller & Jankovics 1984), making them an adequate reference for our work. All the spectra presented in this paper, as well as all the velocities quoted, are in the reference frame of the Na I D interstellar lines.

In addition to these spectra obtained during the MUSICOS 96 campaign, we also used for this analysis two series of spectra obtained previously:

  • in December 1991 and January 1992, with the MUSICOS spectrograph mounted on the 2m TBL telescope at Pic du Midi Observatory, France. These spectra were discussed by Catala et al. (1993). In that paper, the emission lines of Fe II and He I were studied, but we come back to them in order to analyze their photospheric lines. These data were reduced with the MUSBIC reduction software described in Baudrand & Böhm (1992).

  • in November 1994, with the Elodie spectrograph on the 1.93m OHP telescope. These spectra were used in the analysis of short-term photospheric variations of AB Aur (Catala et al. 1997). We re-analyzed them here with the Least-Square Deconvolution method, for an easier comparison with spectra of the MUSICOS 96 campaign.

The log of these previous observations is given in Table 3.


Table 3. Log of the observations: 1991-1992-1994

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: April 28, 1999