Astron. Astrophys. 363, 1019-1025 (2000)

3. Results

3.1. Photometry

In Table 1 photometric data for WDS 00550+2338 obtained during both observing runs are given from which the mean values of apparent brightness in B () and V () bands as well as mean color index B-V = can easily be derived. They coincide with, or are very close to, those reported by Argue (1966), Lee (1970) and Jennens & Helfer (1975). Light curves of WDS 00550+2338 in B, V bands and behavior of B-V color index are shown in Fig. 1

A strong brightness increase on November 27, 1997 is clearly seen in Fig. 1 along with several weaker fluctuations occurring on different dates.

3.2. Spectral data

It is well known that the Harvard MKK system uses for classification of G-K type subdivisions progressive weakening of the ultraviolet part of the spectrum as well as strong G-band and metallic lines at G2-G5 and the appearance of TiO and MgH bands at K5. Given the spectral resolution of our spectrum and taking into account the spectral range, the most representative lines for classification purposes are , , CN band at 4200 A (all wavelengths in continuation are given in A), Ca I 4227, G-band, several strong Fe I lines (4045, 4144, 4271, 4325, 4384) etc.

Since our spectrum is not suitable for precise equivalent width calculation, we used rough estimations for the above mentioned lines to establish the spectral type, comparing our estimated values with those given by Jaschek & Jaschek (1987; 1995) for the same spectral type stars.

To assign the luminosity class, Balmer lines were used since their positive luminosity effect in G-K type stars (Jaschek & Jaschek 1987) is known, for example /Fe I 4071 and Sr II 4216/Fe I 4271 ratios as well as CN strength. On the other hand, Ca I 4227 has a negative luminosity effect amongst late G-K type stars.

With these criteria in mind, using a comparison with the Atlas of spectral tracings of Goy et al. (1995) and Atlas of representative stellar spectra of Yamashita et al. (1977) we conclude that the composite spectrum of 00550+2338 can be classified as K0 IV because:

1. The G-band is the strongest feature (except Ca II K line), even in comparison with a strong Ca I 4227 (which is very strong in the early K type) line, suggesting rather very early K type than late G.

2. The HI lines are still relatively well distinguishable, suggesting either late G or early K subtypes. The strongest lines are Ca II H and K (despite only one of them being seen), G-band, Ca I 4227 (which is stronger than G-band, indicating rather very early K than G type) and several Fe I lines are still relatively weak.

3. The relatively strong G-band is far from being dissolved (as in late K subtypes), an evidence in favor of very early K type.

4. There is no clear evidence for any strong molecular bands which would be indication of a late K type star.

However, the strength of Ca I 4227, several Fe I lines and weak CN band suggest certain characteristics of the K type seen in the composite spectrum. It is therefore resonable to suppose that one of the components belongs to G type while the other shows K type features.

It must be remembered that we are dealing with a composite spectrum which in reality represent none of the components separately. With the knowledge of its binary nature in mind, the spectrum may well be a mixture of an early G (very strong G-band) and not late K (no later than K5 when TiO and MgH bands appear) stars. Taking into account the brightness difference between components (), the spectral types may differ even in the narrower range, namely G5-G8 and K3-K5.

As regards its luminosity class, there are clear indications of the subgiant branch suggested by above mentioned luminosity sensitive lines. At the same time, a comparison with both spectra of luminosity class III giants and normal dwarfs (Goy et al. 1995; Jacoby et al. 1984) shows that WDS 00550+2338 does not belong to any of these classes.

Such a conclusion is well supported by the speckle spectroscopy data showing rather different spectral characteristics of the components (see Fig. 3). The spatially resolved line and the clearly seen Li I doublet at 6708 A first noticed in the composite spectrum by Luck & Challener (1995) suggest that the main component is a young G5-G8 star while the secondary has an almost unseen and may correspond to the K1-K3 type.

Thus, speckle spectroscopy data not only confirm the presence of the Li I doublet but also allow us to attribute that feature clearly to the WDS 00550+2338 A.

3.3. Speckle measurements and orbital data

WDS 00550+2338 has now completed an entire period since its discovery in 1832 by W. Struve, and a number of orbits have been calculated for this bright pair. The most recent calculated orbit, and the one best fitting the observations, has a period P = 167.1 yrs. and semimajor axis a = .002 (Docobo & Costa 1990; see references therein for previous orbits). A number of measurements have been performed since this orbit was calculated, allowing evaluation of its quality. In Table 2, the date of observation, position angle, angular separation, number of observations, name of observer(s), reference source and the residuals of observations made since the last orbit was calculated, with respect to above mentioned orbit, are given.

Table 2. Recent astrometric observations and residuals

From these data, one can see that the orbit is rather well determined and, probably only small improvements would be needed to obtain it definitively.

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk.link@springer.de