Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 362, 673-682 (2000)

Previous Section Next Section Title Page Table of Contents

2. Observations

2.1. Spectra and radial velocities

Eighteen spectroscopic observations were obtained with the 0.9-m coudé feed telescope at Kitt Peak National Observatory (KPNO) during Mar. 31 - April 21, 1998. We used the Ford F3KB CCD with grating A, camera 5, the blue corrector, and the long collimator. The spectra cover the wavelength region between 424 and 454 nm. The resolving power, [FORMULA], was 22,000 corresponding to an effective wavelength resolution of 0.20 Å. The instrumental FWHM was sampled by 2.8 pixels according to a slit width of 400 µm. All spectra were obtained with an integration time of 30 min and have S/N ratios in the continuum between (60-100):1. Data reduction was performed with IRAF and consisted of bias subtraction, flat fielding, and optimized aperture extraction. A representative spectrum near quadrature is shown in Fig. 1.

[FIGURE] Fig. 1. A representative spectrum of BF Aurigae (top). The markers indicate the primary (full lines) and the secondary (dashed lines). As a comparison, we also show a spectrum of the single star [FORMULA] Leo (B5V) and a synthetic spectrum from a 15,000 K and [FORMULA]=4 model atmosphere. The strong line at 4340 Å is H[FORMULA].

Spectra of the radial-velocity standard [FORMULA] Gem (K0III) and the B5V reference star [FORMULA] Leo were obtained at least once during each night to enable an accurate wavelength calibration. However, fitting two Gaussians for the two components to the nightly cross-correlation functions, obtained with IRAF's fxcor task, did not result in the desired small residuals for an individual radial-velocity measurement because, firstly, [FORMULA] Gem is of significantly different spectral type compared to BF Aur and, secondly, the spectral lines of [FORMULA] Leo are too broad to result in a sharp cross-correlation peak. Therefore, we computed a theoretical spectrum from a 15,000 K and [FORMULA]=4.0 ATLAS-9 model atmosphere and used this spectrum as a reference spectrum (Fig. 1). The heliocentric radial velocities of the individual stellar components were then obtained by a least-squares fit of a combination of two of these theoretical spectra, appropriately rotationally broadened, wavelength shifted, and intensity weighted to match the BF Aur spectra. The resulting velocities are given in Table 1 (in column [FORMULA]) and are based on the wavelength shifts of the following spectral features: Balmer H[FORMULA], He ii 438.7 nm, He i 447.1 nm, and Mg ii 448.1 nm. This procedure does not allow to compute a formal error because the overall [FORMULA] also depends on the match of the line intensities rather than solely on wavelength position. We estimate the internal precision of a single measurement to [FORMULA] km s-1 at quadrature and to [FORMULA] km s-1 near conjunction based on a comparison with other data taken during the same nights.


Table 1. Radial velocities in km s-1 (P = primary, S = secondary). The last two columns, [FORMULA], denote the heliocentric radial velocities of the primary and secondary. Phase has been computed with the ephemeris given in Demircan et al. (1997).

2.2. Photometry

Johnson UBV photometry was obtained with one of the two 0.75-m Vienna Observatory automatic photoelectric telescopes in southern Arizona (for details see Strassmeier et al. 1997) in the time between January and March 1998 just prior to the spectroscopic observations.

Altogether, 387 U-, 342 B-, and 320 V- points are presented in Tables A1-A5 (only in electronic form). The integration time was 60 s per reading and the observations were arranged in the sequence C2 - S - C1 - V - C1 - S - C2, where V denotes the measurement on BF Aur, C1 refers to the comparison star, HD 32330 (B2IV), C2 to the check star, HD 32418 (A4V), and S to sky measurements. This sequence was carried out repeatedly for U, B, and V, and took 7 minutes per filter including recentering the star after each filter sequence. For the reduction procedure, we use the average [FORMULA]V-C1[FORMULA] count rates of each sequence. The standard error of a single C2-C1 mean from the overall mean was [FORMULA] in BV and [FORMULA] in U.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: October 24, 2000