Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 347, 583-589 (1999)

Previous Section Next Section Title Page Table of Contents

3. Several variabililty states?

Fig. 4 presents the skewness calculated for the line NIV [FORMULA] present in the IUE wavelength range, plotted as a function of phase computed with the new ephemeris. Fig. 4a contains data obtained in 1983 and 1995, while Fig. 4b contains data obtained in 1988 and 1992, and although each panel of this figure displays a regular behavior of the skewness with phase, there is a clear difference in the number of maxima present. The top panel (Fig. 4a), which contains the data obtained in the IUE MEGA Campaign, display 3 maxima per cycle, while the bottom panel (Fig. 4b) displays only one. Thus, we conclude that there are at least two variability states in the system which are recurrent, one "active" state in which more variability is present per cycle, and a second, more "passive" state. The remarkable aspect of this shift from one state to another is that each state retains coherence in the variability, despite being separated by many cycles from a previous, similar state.

[FIGURE] Fig. 4a and b. Skewness of the N IV 1718 line as a function of phase with the same ephemeris as in Fig. 1. a  "Active" state illustrated by data obtained in 1983 (open circles) and in 1995 (filled circles, IUE MEGA campaign). b  "Passive" state illustrated by data obtained in 1988 (filled-in diamonds) and in 1992 (open squares).

The photometric variability of EZ CMa has a behavior similar to that of the skewness in that the light curve is variable in shape and amplitude (Duijsens et al., 1996). However, here also one can identify patterns in the shape of the light curves which recur, and one can thus extend the concept of different variability "states" to the photometric behavior. For example, the data of Cherepashchuk(1981) and Robert et al.(1992), obtained respectively in 1980 and 1990 have almost identical phase-dependent behaviors, as can be seen in Fig. 5a. These light curves have a maximum near phase 0.5 and a trend towards a maximum at phase 0.0 which is broken by a minimum. The other two light curves plotted in Fig. 5b and 5c (data obtained in 1988 (Robert et al., 1992) and 1987 (Drissen et al., 1989), respectively) have maxima near phase 0.0, with only a hint of a maximum near phase 0.5 A similar story is illustrated in Fig. 6: the data of Balona et al.(1989a), obtained in 1986, and Lamontagne et al.(1986), obtained in 1985, have almost identical shapes as illustrated in Fig. 6c, with a prominent maximum near phase 0.65 and a hint of a maximum near phase 0.2. The other light curves in this figure, corresponding to data obtained in 1993 (Antokhin et al., 1995) (Fig. 6a) and 1990/91 (Duijsens et al., 1996) (Fig. 6b) also have maxima near these same phases, although the maximum near phase 0.2 is much more pronounced. We emphasize that the phases computed for the data plotted in Figs. 5 and 6 have the same initial epoch and period as all the rest of the data in this paper, and no artificial phase shift has been introduced to match up the data. Hence, once again, it is remarkable that when the light curves are very similar, they display coherent variations (as in Figs. 5a and 6c), despite being separated by numerous cycles. Furthermore, it is interesting to note that there are two sets of phase intervals within which a maximum can occur: (0.2, 0.65), "state" A; and (0.0, 0.5), "state" B. The photometric data used are described in Table 2.

[FIGURE] Fig. 5a-c. Light curves plotted with the same ephemeris as in Fig. 1 for the sets of data summarized in Table 2. The plotted data were obtained in the years: a  1980 (filled-in circles) and 1990 (open circles); b  1988; c  1987. These light curves all display maxima near phases 0.0 and/or 0.5.

[FIGURE] Fig. 6a-c. Same as Fig. 5 but for different sets of data, obtained in the years: a  1993; b  1990/91; and c  1986 (open circles) and 1985 (filled-in circles). These light curves all display maxima near phases 0.2 and/or 0.65.


Table 2. List of the photometric data.
a) See text for details
b) Shifted by 0.015 mag to fit the data of Robert et al. (1992)
c) Shifted by 0.820 mag to fit the rest of the data
(1) Cherepashchuk 1981; (2) Lamontagne et al. 1986; (3) Balona et al. 1989b; (4) Drissen et al. 1989; (5) Robert et al. 1992; (6) Duijsens et al. 1996; (7) Antokhin et al. 1994;

What is responsible for the maxima in the light curves? What physical process can switch the system from one state to another one, and yet retain coherent variability within each of the two different states? It is beyond the scope of this paper to even attempt providing aswers to these questions. All we are able to conclude at this time is that the phase coherence of the data, separated by many years, points to a common underlying periodic process with a stable ephemeris.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1999

Online publication: June 30, 1999