The HRI in-orbit background is composed of several components:
a) The internal background is believed to come mainly from residual radioactivity in the MCP glass and the surrounding detector structure. This was measured on the ground with a rate of 1.5 counts s-1 over the entire device. This appears to be a stable component of the background and is uniform over the detector.
b) The X-ray background (XRB) contribution (comprised of both galactic and extragalactic components) varies across the sky, with a typical value in the ROSAT energy band of 1.0 count s-1, over the detector.
c) The externally induced background from charged particles, earth scattered solar radiation, and auroral photons. On average, the externally induced background rates are higher for ROSAT than for Einstein due to the higher inclination and altitude of the orbit. The external background rates vary considerably during an orbit and also from orbit to orbit, with a minimum of 1 count s-1 up to about 10 counts s-1, at which point the HRI high voltage is reduced to protect the detector from overload.
Figure 16 shows the variation in the background rate during an HRI observation of the North Ecliptic Pole (NEP). The lowest rates occur when the satellite is crossing the geomagnetic equator and the highest rates occur at the northern and southern extremities of the orbit. In many cases the HRI high voltage needs to be reduced in these regions to protect the detector from overload. Table 5 summarizes the average in-orbit values (and ranges) of all the background components.
|Internal||1.5 (1.0-2.0)||counts s-1|
|External||3.0 (1.0-10)||counts s-1|
|XRB||1.0 (0.5-2.0)||counts s-1|
|Total||5.5 (2.5-14)||counts s-1|
|7.8 x 10-1||counts s-1 cm-2|
|3.8 x 10-3||counts s-1 arcmin-2|
|1.1 x 10-6||counts s-1 arcsec-2|
Figure 17 compares the pulse height analyzer (PHA) distribution of the background in the NEP observation during low background intervals (count rate < 5.0 counts s-1; dominated by detector noise and the XRB) and high background intervals (count rate > 5.0 counts s-1; dominated by externally induced background events). Even though the background is dominated by different particles during these intervals, the spectra are consistent given the very modest energy resolution of the HRI. The background rate is the greatest in the lowest PHA channels since radioactive decay in the MCP glass and charged particles can produce electrons anywhere within the MCPs. These electrons are accelerated through through a smaller potential difference than electrons produced near the entrance of the MCPs. The background in PHA channels < 10 is very uniform across the detector. The background in PHA channels > 10 is enhanced beyond 15' off-axis due to the greater gain in the detector (see §3.5).
Also shown in Figure 17 is the PHA distribution in an HRI observation of AR Lac. Notice that the AR Lac spectrum is much more strongly peaked than the background spectrum. Most of the AR Lac source photons fall between PHA channels 3 and 7. In general, to obtain maximum sensitivity for detecting point sources or for analyzing low-surface brightness extended objects, the user is advised to screen out the highest background intervals and only examine PHA channels between approximately 3 and 8, inclusive. The optimum range of PHA channels depends on the source, the source position on the detector, and the date of observation. The user is advised to compare the source and background spectra for each observation to determine which channels should be screened out.