Next: The near infrared range Up: The optical range Previous: The orbital inclination
Oscillations
In order to check if a coherent oscillation hides in the flickering we calculated Lomb-Scargle periodograms ([Lomb 1976], [Scargle 1982], [Horne & Baliunas 1986]) of the light curves. In order to reduce the low frequency signals introduced by the orbital modulations the eclipses were first cut out, and a spline fit to a binned version of the light curves (bin width: 15 min) was subtracted.
A significant signal leading to a convincing curve when folding the light curve on the corresponding period was found on 1999 June 21/22. The periodogram is shown in the upper frame of Fig. 5. The peak at 2.92 +- 0.01 x 10-3 Hz suggests a period of Posc = 5.72 +- 0.02 min. The light curve, folded on this period, is shown in the central frame of Fig. 5. Although the random flickering dominates the curve exhibits definitely a sinusoidal modulation. This is better seen in the lower frame of Fig. 5 which contains a binned version (bin width: 0.01 x Posc ) of the light curve in the central frame. The solid line represents the best fit sine curve which has a half-amplitude of 0.0319 +- 0.0007 mag .
The light curves of the other nights do not exhibit a signal at the corresponding frequency (or a convincing modulation at other frequencies). Only the periodogram of 1999 May 9 contains a conspicious peak (although much less significant than on 1999 June 21) at a nearby frequency of 2.72 +- 0.01 x 10-3 Hz. Folding the light curve on the corresponding period (6.14 +- 0.02 min) results in a curve with a narrow minimum and a double peaked maximum, i.e. far less sinusoidal than in the previous case. However, the total amplitude of the variations is similar in the two nights.
The periodogram of the combined data of the four nights of high-speed photometry, shown in the insert in the upper frame of Fig. 5, contains two maxima - each one consisting of many alias peaks - close the the frequencies of the signals found on May 9 and June 21/22, but slightly shifted to higher and lower frequencies, respectively. The distance between the centroids of these maxima corresponds well to the orbital frequency of V893 Sco, suggesting them to be a periodic signal and its beat frequency with the orbital cycle.
The modulations discussed above are reminiscent of the variations seen in intermediate polars, although the present observations do not permit to regard them as stable oscillations of V893 Sco. However, their presence is intriguing, and future observations should be scrutinized for similar signals.