2. Observations and data reduction
More than two hundred spectra have been collected since 1992 at Ondejov, Haute Provence and Dominion Astrophysical Observatories (OND, OHP and DAO hereafter). The basic characteristics of the instrumentation used are summarized in Table 1 and the actual numbers of spectrograms secured with individual instruments over the years 1992 - 1999 are in Table 2. Part of the data secured at OHP and DAO were obtained as nightly series of spectra. More details on these series are given in Table 3. This table also provides information on the typical exposure times at OHP and DAO. The actual exposure times varied broadly from 1200 to 3600 s for the OHP spectra, from 250 to 1560 s for the spectra taken at DAO, and from 1758 to 18300 s for OND. The signal-to-noise ratio of most of the spectra was between 300 and 400. The spectra obtained from 1992 to 1994 at OHP and all OND spectra were reduced with the software SPEFO written by the late Dr. J. Horn (Horn et al. 1992, Skoda 1996). The reduction of the data obtained at OHP in 1996 was made with the code written by JC. Observations from DAO were reduced partly with IRAF (basic extraction, flatfielding and creating 1-D images) and partly with SPEFO (wavelength calibration, rectification, RV and intensity measurements).
Table 2. Number of spectrograms of 60 Cyg obtained each year with the individual instruments defined by their codes in Table 1.
Our data analysis was concentrated on several spectral lines of H I and He I for which enough data was accumulated. We performed frequency analysis of data describing the profile variations and also analysed all basic quantities such as RV, central intensity, equivalent width (EW hereafter) and FWHM, and also the peak intensities of the double H and He I 6678 Å emission whenever possible.
The study of the profile variations was made by Fourier analysis and the CLEAN algorithm which applied to time series of spectra obtained in the blue region and are listed in Table 3. Least-squares sinusoid fitting was applied to the same time series. These are standard approaches based on search for periodicity in the variations in normalized intensity as a function of position in the line profile. For three He I lines, at 4388, 4471 and 6678 Å, we used a Fortran program HEC26 which measures local RVs (in fact, position on the profile transformed to RV via given wavelength) at specified line depths, as suggested by Harmanec (1999) and the results were analysed with period search codes. This new method is, in a sense, complementary to the CLEAN technique, but the the long-term variations of the radial velocity can be easily taken into the account. In contrast to it, the existing applications of the CLEAN method usually assumed a constant RV of the studied object. Program HEC26 (written by PH) uses digital spectra in a tabular form wavelength vs. relative intensity and derives a table of local RVs interpolated for pre-determined line intensities by a linear interpolation. It is also possible to smooth the profiles in the process to suppress the noise. This procedure has been advocated for some time by one of us (cf. Harmanec 1999) but this study represents the very first practical application of it.
The analysis of the above mentioned basic quantities was performed with various period search codes. The definition of these quantities is obvious with the exception of the radial velocity, namely in the case of asymmetrical profiles. Several methods, including the HEC26 code were applied. For their description, limitation and physical meaning of particular measurements in the case of early-type stars see Harmanec (1998a).
All measured quantities are presented in Table 4. 1
This study is based on photoelectric photometry of 60 Cyg from five observatories:
The first three sets of differential observations were obtained relative to HD 199311 and the check star HD 199479 was also regularly observed. Since both these stars are much fainter than 60 Cyg, we have chosen another comparison, HR 8161, a B0.5 V star, whose constancy was confirmed by the Hipparcos data. This new comparison was used during the 1998-99 observations at SPM and Hvar. Both original comparisons were also observed as the check stars in 1998. HD 199311 served as the only check in 1999.
Xinglong, Hvar and SPM data were transformed into the standard UBV magnitudes with the help of non-linear transformation formulæ using Fortran program HEC22 (Harmanec & Horn 1998, Harmanec et al. 1994). The Toronto BV observations were transformed into the standard Johnson magnitudes with the help of a linear transformation by Percy et al. (1997). To obtain the standard UBV magnitudes of 60 Cyg, the following improved all-sky UBV values for HD 199311, derived by Harmanec et al. (1994), were added to the respective magnitude differences 60 Cyg - HD 199311:
V = 6690, = 0080, = 0098
for all observations secured prior to 1998. The mean transformed Hipparcos V magnitude of HD 199311, based on Harmanec's (1998b) formula, is 6685 0008. All photometric observations used in this study are stored in the Ondejov photometric data archives and are available on request from PH (hecsunstel.asu.cas.cz).
The reduction of 1998 data proceeded in two steps. First, we obtained improved all-sky UBV values of HR 8161 using observations of many Hvar standards (see Harmanec et al. 1994) on good nights at SPM and Hvar. Then, we adopted these improved values
V = 5764, = -0126, = -0515
and reduced the whole season at SPM and Hvar with them. For comparison, the mean transformed Hipparcos V magnitude of HR 8161, based on Harmanec's (1998b) formula, is 5764 0005.
To illustrate to which extent we achieved the homogeneity of all four photometric data sets, we summarize in Table 5 the mean magnitudes of the check star HD 199479 from all four stations. The check star has also been observed by the AAVSO observers in another program (Percy et al. 1997). We also list the mean value from this additional source. For 1998, 1999 and Hipparcos data, we also quote the mean values for HD 199311, now observed as another check star relative to HR 8161.
Table 5. Mean UBV values of the comparison and check stars used: All quoted errors are the rms errors of 1 observation of unit weight
We also compiled UBV observations of 60 Cyg from the astronomical literature. Their number is very limited and most of such observations are published without the dates of observations. We summarize these data together with estimated Julian dates of observations in Table 6.
Table 6. Published UBV photometry of 60 Cyg.
© European Southern Observatory (ESO) 2000
Online publication: April 17, 2000