Spacecraft pointing jitter causes variations in the scattered light and in the measured stellar flux (due to residual errors in the flat fielding of the CCD). The noise power due to scattered light variations depends on the design details of the telescope and will be evaluated during the assessment phase. The noise power due to residual flat fielding errors may be estimated by
| (4) |
The symbols have the following meaning: n= number of pixels in the defocused image; residual relative pixel-to-pixel variation in the responsivity of the CCD; angular size of one pixel; r.m.s. pointing fluctuations which have a characteristic time scale . We may assume the following typical numbers: n=100, ,, and s. With a good calibration a residual flat fielding error of 10-3 should be achievable. In this way we find that ppmHz-1, which is of the order of the photon noise of an solar-type star. This example illustrates that this noise component need not be a major source of concern.
There are, finally, many slow drifts and slowly varying parameters affecting the photometry at low frequencies. Let us consider a quasi periodic variation with a typical time scale . Without further details it is not possible to evaluate the absolute noise levels; these will be studied during the assessment phase. There is, however, a general consideration which illustrates that the influence at high frequencies will be very small. This is based on the following estimate of the relative noise levels at high frequencies and low frequencies
| (5) |
Taking for example hr and mHz, we find that . Hence, although the noise power at low frequencies () may be many orders of magnitude above the photon noise, things can be arranged so that the noise power at the frequencies of interest () is totally negligible. This is due to the fact that we employ differential photometry.