Astron. Astrophys. 322, 511-522 (1997)

4. The photometric variations

4.1. Periodogram analysis

In this paper we use the multi-periodic search program PERIOD (Breger 1990) that allows us to fit and improve multiple frequencies simultaneously without prewhitening. Our periodogram analysis of all combined V and y -band observations from 1996 (see Fig. 1a) had the strongest peak at 5.913 0.005 days ( d^-1). The maximum amplitude was slightly over 0.11 mag (Fig. 1e). Additionally, we see a series of weaker peaks at the aliasing frequencies (), with the =0.8309 d^-1 = 1.2035-day period being the second strongest as expected. The prewhitened data with in Fig. 1f demonstrate the aliasing character of the 1.2-day period and just show the one-day aliases. Once again, we conclude that the 5.9-day period is the true photometric period and the 1.2-day period its alias. This result is in very good agreement with previous results based on different data and different analyses (Noskova 1989, Bond & Livio 1990, Kuczawska & Mikolajewski 1993, Hubl & Strassmeier 1995, Strassmeier et al. 1997).

Malasan et al. (1989) pointed out that phase shifts of the minimum light of +0.11 and -0.37 occurred when combining their data from 1988 and 1989, respectively with those from Schnell & Purgathofer (1983) from 1983 and phased with days. If phased with the 5.9-day period we do not see such phase shifts in our data (Fig. 2). Notice that the large dispersion in Fig. 2 is not due to instrumental scatter but to instrinsic variations of the stellar brightness.

Fig. 2. The entire and photometry from Fig. 1a phased with the best seasonal photometric period (5.913 days and Eq. 1). The color of the star is redder at light-curve minimum, which is expected if the spotted regions are cooler than the photosphere. Despite the long baseline of observations (February through July 1996), the phase of the light-curve minimum remains stable. Notice though that the "scatter" is not instrumental but due to intrinsic variations of the star.

In this paper we chose to phase all data with the well defined photometric period from the present data

which we interpret to be the rotation period of the G5 star, and the initial epoch is a time of one of the first light-curve minima in that data set.

4.2. Average brightness and colors

The average V brightness and broad-band colors remained basically unchanged from 1993 to 1996, and agree with those previously observed. However, as already noted by Jasniewicz et al. (1994), the average V value of 8.66 0.02 given by Malasan et al. (1989) for 1989 was brighter by over 0.2 mag and their of +0.78 0.03 bluer by 0.05 mag. Fig. 1b indicates an increasing reddening of by approximately 0.010 mag between February and July 1996 possibly due to a spot cycle.

4.3. The rotation period

Given the 5.9 or the 1.2-day period as the only observationally constrained values that could be interpreted as the rotation period of IN Comae, we may argue that a 1.2-day period would lead to an unacceptable high rotational velocity and that therefore the 5.9-day period is also most likely to be the rotation period of the G5 star.

This conclusion is based on our new and precise value for the projected equatorial rotational velocity ( 1 km s^-1 ; see Sect. 6 and 7) and the G5III-IV spectral classification. Schmidt-Kaler (1982) lists a stellar radius for a G5III star of 10  , in agreement with other sources. If we adopt a radius between 5 and 10  for a G5III-IV star, a rotation period of 1.2 days would mean an equatorial rotational velocity between 210 and 417 km s^-1. As a comparison, the most rapidly-rotating late-type giant star known to date is FK Comae with a of 159 4 km s^-1 (Rucinski 1990). Adopting its inclination angle of obtained from Doppler imaging by Piskunov et al. (1994), an unprojected equatorial velocity for FK Comae of 165 km s^-1 would result. Thus even at the lower end of the range of values for the equatorial velocity of IN Comae implied by a period of 1.2 days, IN Comae would be the most rapid rotator among the late-type stars and further, at velocities near 210 km s^-1, the inclination angle of the stellar rotation axis would have to be a mere to match the observed .

On the other hand, with the 5.9-day period the measured puts a lower limit to the stellar radius of 7.8 0.2  . If we adopt as a reasonable lower limit for the detection of significant rotational modulation due to surface features - as observed -, an upper limit for the radius of 15  would be still consistent with the measured . It follows that only the 5.9-day period is a reasonable rotation period for this G5 giant and that an adopted inclination of also agrees with a "typical" stellar radius for a G5III-IV star, making IN Comae almost a twin to the FK-Comae-type star HD 199178 (Strassmeier et al. 1996b).

4.4. The 0.25-day period

Fig. 3 shows the differential B and V light curves of IN Comae during two nights in April 1983 (Schnell 1996). It seems obvious that there was no 0.25-day period at that time but instead the data showed the expected trend for a sinusoidal 5.9-day period. Furthermore, our new photometry from 1996 with up to ten points per night also did not show any periodicity except the 5.9-day period and its aliases. Therefore, the 0.25-day period found by Kuczawska & Mikolajewski (1993) is either a time-variable phenomenon in IN Comae or their comparison star SAO 82575 is a variable with that period.

Fig. 3. High time resolution and monitoring on April 19, 1983 (left panels) and April 25, 1983 (right panels) by Schnell (1996) does not show a 0.25-day periodicity as found by Kuczawska & Mikolajewski (1993) in 1993.

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de