Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 334, 606-608 (1998)

Previous Section Next Section Title Page Table of Contents

3. Frequency analysis

The DFT of the combined data in Table 1 was computed in the range [FORMULA]  mHz. This revealed that all oscillatory power is confined to the vicinity of 1.4 mHz and that no harmonics are present. Fig. 3 shows the DFT of the combined data in the narrower range 1-2 mHz. Owing to the short light curves, the spectral window of these data is very strongly aliased. The dominant window pattern in the top panel of Fig. 3 is centered at [FORMULA] =1452 µHz. We fitted this frequency to the data by linear least squares to establish its amplitude and phase, with their respective fitting errors. The results of this fit were then used as initial values for an iterative non-linear least squares fit which optimizes frequency, amplitude and phase together. A sinusoid with these optimized parameters was then subtracted from the data and the DFT of the residuals was calculated to produce the middle panel in Fig. 3. This suggests the presence of another frequency whose window pattern is centered at [FORMULA] =1411 µHz. The two-stage least squares fitting procedure was repeated for both frequencies together (Table 2) and they were subtracted from the data to produce the residuals shown in the bottom panel of Fig. 3. This shows no further signals above the noise level of 0.2 mmag.

[FIGURE] Fig. 3. (Top) The amplitude spectrum of the combined JD 2 450 655-663 data showing a dominant oscillation at [FORMULA] =1452 µHz. (Middle) Residuals after prewhitening [FORMULA] reveal another oscillation at [FORMULA] =1411 µHz. (Bottom) The amplitude spectrum of the residuals after removal of [FORMULA] and [FORMULA]. The ordinate scale is identical for all three panels.


Table 2. The frequency solution for HD 213637.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: May 15, 1998