3. Fourier analysis
A period search on the "cleaned" datasets (SAA and stray light corrected) was carried out with a standard Fourier technique (Deeming 1975). The resulting amplitude spectra for both stars reveal a number of significant frequencies (see Tables 2 and 3). A signal is considered to be significant, if the amplitude is at least four times larger than the noise level (Paper I). Since stars in the continuous viewing zone were observed, the duty cycle was large and hence the window function of the Fourier transform is clean with very small side lobes.
Two groups of frequencies were determined and heuristically identified. One group, which seems to be related to the orbit () and another one, of yet unknown source, which is clearly present in both guide star data sets and hence is probably not intrinsic but of instrumental origin. The orbital modulation which remains after correction for the SAA comes from the residual background modulation (Fig. 2) and the averaging procedure over 24h when determining the stray light correction.
Table 2. List of eliminated frequencies for GS-54 (1 d-1 = 11.6 µHz).
Table 3. List of eliminated frequencies for GS-75
In the next step the light curves of both guide stars were prewhitened with the frequencies and amplitudes determined with a simultaneous multi-sine fit, as given in Tables 2 and 3. The amplitude spectra of GS-54 and GS-75, cleaned for identified non-intrinsic effects, are shown in Fig. 4. In the low frequency domain from 0 to about 0.1 mHz for GS-54 (noise level at 23 ppm) and from 0 to 0.2 mHz for GS-75 (noise level at 47 ppm) numerous peaks can still be found above the 99.9% significance level (dotted lines). Whether these peaks are caused by long term stellar variability or can be interpreted as signal due to granulation similar to our sun, or by complex instrumental effects cannot be decided at this time.
© European Southern Observatory (ESO) 1999
Online publication: March 1, 1999