The time variability of X-ray emission can be studied with the HRI since each detected photon has its time of arrival recorded by the detector. The accuracy of this time is limited by the electronic resolution of the HRI processor which is 61 s relative to the ROSAT spacecraft clock. The relative arrival times of photons during a single observation is accurate to this value. The absolute accuracy of the ROSAT spacecraft clock, and its conversion to UT, is expected to be a few milliseconds.
The HRI has a processing dead time during which events may not be counted which varies between 0.36 and 1.35 ms per event. The variation is discussed in Section 2.7 and depends on the fine position of the event. Thus there is a dead time correction that needs to be made for calculating the true event rate from a source. A mean dead time of = 0.81 ms can be used for this purpose, and the true rate is then given by: nt = n(1 + n), where n is the observed rate.
Due to the telescope wobble and the small variations in the QE of the HRI on spatial scales of a few arcminutes (see Figure 15), the count rate of a source can vary by 5% between the extremities of a wobble. This can produce a low amplitude source variability on a time scale of approximately 100-400 s in some sources.
As for the PSPC, the HRI observations are typically interrupted once per orbit, and sometimes as much a three times per orbit. Typical continuous viewing times for a source will be about 2000 s, with some cases lasting up to 4000 s. Long term monitoring of sources on time scales of weeks of more will be limited by the solar view constraints of the satellite. This limits source accessibility to about one month every six months for a source in the elliptic plane, with greater access time for sources closer to the ecliptic poles.