2. Observations and reduction
Table 1 contains information about all the photometric, spectroscopic and magnetic data that we used in our period analysis. We divided all data by years; the JD range of each observational run is in the second column of Table 1. The number of observations is given in the fourth column. We present below our new observations together with a short description of the previous observational data.
Table 1. Observational data sets used in the paper
The photometric observations represent the most extensive data sets. The new Strömgren four-color photometric data were obtained with the Four College Automatic Photometric Telescope (FCAPT) in 1991-1997. Both the telescope and the reduction procedure are discussed by Pyper et al. (1993). The data are too extensive to be printed, so they are given in Table 5 in electronic form.
From previously published photometric observations only one set made in 1964-66 by Abuladze (1968) was not included because of extremely large dispersion of the observational points. From uv-set #7 (hereafter "set #" refers to Table 1), we used only the photometry in the 3330 Å spectral band which is close to the U and u bands.
There is a good set of uvby photometric observations made in May 1974 (Weiss et al. 1976), but their data are not available as they are published only as plots.
2.2. Equivalent widths and radial velocity measurements
New spectroscopic observations of CU Vir were made in June 1994 and in March - May 1997 at the coudé spectrograph of the 2.6 m telescope of the Crimean Astrophysical Observatory with the CCD detector attached. Some of the observations were made with the Zeeman analyzer using the Stokesmeter (Plachinda et al. 1993). The observed spectral region, 6325 - 6385 Å contains two strong Si II lines. All spectra were taken with a linear reciprocal dispersion of 2.5 Å mm-1 and corresponding spectral resolution of about 0.2 Å. The signal-to-noise ratio were in the range 150 - 300. Magnetic measurements will be described below.
The reduction of the spectra was made using the software "SPE" written by S. Sergeev at the Crimean Observatory. The reduction procedure includes the night sky subtraction, flat field correction, normalization of spectra to the continuum, cosmic ray subtraction by visual inspection of the spectra and wavelength calibration.
The heliocentric Julian dates of the midpoints of the exposures, and equivalent widths of the Si II 6347 line are given in Table 2.
Table 2. Journal of spectroscopic observations of CU Vir made at the Crimean Astrophysical Observatory in 1994-1997.
The previous spectroscopic observations consist of equivalent widths and radial velocity measurements. The latter were measured for hydrogen lines by Abt & Snowden (1973) and by us using spectra obtained in 1994 at Observatoire de Haute Provence (kuschnig et al, in prep.). Radial velocities were measured for the center of gravity of the line core and are given in Table 3 together with the heliocentric Julian dates. The typical rms is about 1 km s-1.
Table 3. Radial velocities, measured by line.
For the period analysis we used the equivalent widths of the Si II lines because they usually vary in phase with the light variations. Equivalent widths and radial velocities of He I and hydrogen lines, and -photometry were used as a final check on the validity of new periods. For observational set #4 only plots of the equivalent widths are published by Krivosheina et al. (1980), but we have all the information in digitized form which is available through e-mail request to Ryabchikova. We also corrected typographic errors in the JD's for set #12.
We excluded the spectroscopic data by Peterson (1966) from our analysis because the small number of points together with the quality of the equivalent widths measurements on photographic plates did not allow us to properly determine the maxima and minima of the spectral variations.
2.3. Magnetic field measurements
The effective magnetic field is a component of the magnetic field vector along the line of sight, averaged over the stellar disk. We have two sets of effective magnetic field measurements separated by 20 years (data sets #8 and #29). The first set was made with a photoelectric polarimeter which measured the circular polarization in the wings of the line (Borra & Landstreet 1980).
New observations of the effective magnetic field were carried out with the Stokesmeter and CCD detector in 1996-1997 using the strong Si II 6347.09 and 6371.36 lines with effective Landé factors 1.167 and 1.333, respectively. The Stokesmeter is mounted in front of the entrance slit of the coudé spectrograph of the 2.6 m telescope. It consists of two rotating achromatic quarterwave plates and a plate of Iceland spar between them to separate right- and left-circular polarization spectra. The typical exposure time for magnetic measurements was about one hour which corresponds to 0.08 of the rotational period.
Synthetic spectrum calculations show that both Si II lines are practically free of blends in the CU Vir spectrum. Unfortunately silicon has a nonuniform distribution on the stellar surface, therefore the effective magnetic field measured with Si II lines may differ from that measured with the hydrogen polarimeter. As was shown by Kuschnig et al. (in prep.), the silicon distribution on the surface of CU Vir does not have a very complex structure. It consists of one depleted spot and a larger zone rich in silicon, so we do not expect there to be significant differences between magnetic measurements made with the Si II lines and the line. We measured a shift between the centers of gravity of the right- and left-circular polarization line profiles and then converted it to the effective magnetic field.
New measurements of the effective magnetic field are presented in Table 4. Each value is the result of averaging both Si II lines. The rms of the magnetic measurements is 185 G.
Table 4. Journal of effective magnetic field observations of CU Vir made at the Crimean Astrophysical Observatory in 1994-1997.
The low accuracy of the magnetic measurements in CU Vir does not permit them to be used in a period search. Since the two data sets are well-separated in time, we mainly consider them as additional support for our period solutions.
© European Southern Observatory (ESO) 1998
Online publication: October 22, 1998