Astron. Astrophys. 322, 493-506 (1997)

2. The data

The data were obtained by P. Mathias and by C. Aerts at the Observatoire de Haute Provence with the Aurélie spectrograph on the 1.52m telescope, and consist of 660 spectra taken on 19 nights over a total time span of 33 days (May 11 - June 14, 1992). The spectra are sampled with a resolving power of R 50000, over a wavelength range which comprises absorption lines of the Si III triplet 4552.6, 4567.8, 4574.8 and of O II 4591.0. The Si III 4552.6 is blended with a weak line of Si III 4554.0. All spectra are mapped to the heliocentric frame and the times of mid-exposure are in HJD. For this study we used only the best 620 spectra (S/N 150). For a detailed discussion on the used instrumentation and on the data we refer to Aerts et al. (1994). Fig. 1 displays the mean spectrum in the regions around the absorption lines. An example of one night of data is shown in Fig. 2.

Fig. 1. Mean of the 620 spectra in the wavelength regions that were used to normalize the spectra

Fig. 2. An example of one night of data, in the wavelength range around the Si III 4574 line. From top to bottom: the mean of all 620 spectra - 54 normalized spectra taken on JD 2448756 - a grey-scale representation of the intensity variations in the line - the 54 residual spectra (mean spectrum subtracted from the normalized ones) - a grey-scale image of the residual intensity variations in the line

In Fig. 3 we present a comparison of the equivalent width (EW) changes of the optical lines with those of the UV wind lines. We plot the EW variations of the Si III 4552 line, and those of the UV resonance doublets of Si IV ( 1394, 1403) and C IV ( 1548, 1551). Since there were no UV data taken simultaneously with the optical data, we represent the UV EW data with an interpolating fit to IUE data presented by Veen et al. (1996). (There are IUE measurements taken less than a year before and less than a year after our optical measurements.) The fitted function consists of two sinusoids with periods of 12.0 and 6.0 days, and is fitted to 75 measurements, spanning 14 years, for C IV and to 67 measurements for Si IV. With this function, all UV EW data points are fitted well within the error of the measurements; a fit with just one sinusoid with a 6 day period does not give such a satisfactory result. The combination of the two sinusoids makes it clear that the EWs of the UV wind lines vary with alternating deep and less-deep minima. For this reason and because of the similarity of the UV wind variations with those of the magnetic star HD184927, Henrichs et al. (1993) and Veen et al. (1996) interpret the UV variations in Cep as due to the consecutive passing of the poles of a magnetic dipole, which is oblique to the rotation axis of the star.

Fig. 3. Variations of the equivalent width of lines in the ultraviolet and in the optical. The top panel shows the interpolated fit to the EW variations of the UV resonance doublets of Si IV (top curve) and C IV (bottom curve). The lower panel shows the EW changes of the Si III 4552 line. Variations with a time scale of 12 or 6 days are superposed on the variations caused by the radial pulsation. The sampling of the optical data is such that we cannot distinguish between deep and less-deep minima, as in the case of the UV resonance lines. Note that with each UV deep minimum we find a minimum in the EW of the optical lines as well

One can clearly see in Fig. 3 that, although the sampling of the optical data is troubled by weather problems, the photospheric profiles also show evidence for a 6 and/or 12 day periodicity.

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de