2 Observations and reductions

CS 31082-001 was first observed at high dispersion (R > 45 000) with the new ESO-VLT spectrograph UVES (Dekker et al. 2000) in August 2000, and was immediately found to display outstandingly strong europium and thorium lines. The star was promptly re-observed at the highest possible resolution (R=75 000 in the blue arm of the spectrograph) during a subsequent night, and again in October 2000. Most of the spectra were obtained simultaneously in both the blue and red arms of UVES, using a dichroic mirror, as summarized in Table 1.

Reductions were performed using the UVES context within MIDAS. The succession of tasks included bias subtraction from all images, fit and subtraction of the inter-order background from the object and flat-field images, and the averaging of ten flat-fields per night into a master frame. The echelle orders of the object were optimally extracted (assuming a Gaussian profile for the object perpendicular to the dispersion, and a constant for the sky background); the same weights were applied to extract the flat-field, which was then used to correct both for pixel-to-pixel variations and the blaze of the instrument. The wavelength calibration was performed on Th-Ar lamp frames and applied to the extracted object. The spectra were finally resampled to a constant wavelength step and normalized to unity by fitting a spline function. All spectra obtained with identical spectrograph settings (same cross-disperser and same central wavelength) were then co-added after radial-velocity correction. The lower part of Table 1 summarizes the obtained signal-to-noise ratio per pixel (1 pixel $\sim$ 0.0015 nm) at wavelengths of interest.

Radial velocities given in the table are only those computed from the 480-580nm red spectra (which permitted corrections for small instrumental shifts from the telluric absorption lines). In addition to the spectra reported here, we took exposures of CS 31082-001 with UVES again in September 2001 (5-9 September) and the radial velocity was still impressively unchanged (mean of four measurements in 2001 $V_{{\rm r\: barycentric}}=139.05\pm0.05~{\rm km~s}^{-1}$) and Aoki (2002, private communication) found $V_{{\rm r\:barycentric}}=138.9\pm0.24~{\rm km~s}^{-1}$ from observations with HDS on the Subaru telescope. There is therefore no sign of radial velocity variations in this star, and hence no hint that CS 31082-001 could be a binary.