XMM-Newton Science Analysis System: User Guide

4.4.2 Filtering EPIC-pn concatenated event lists

The EPIC pn camera consists of twelve 64 x 200 pixel CCDs on a single wafer. For full frame and extended full frame modes the first 12 rows at the readout node are not transmitted to ground (are set to "bad", equivalent to "bad pixels"). In contrast to the MOS, all non bad pn events supposedly related to X-rays are transmitted to the ground and the pattern recognition and recombination of split partner is done off-line by the task epevents. Filtering of an EPIC pn dataset is entirely performed by the EPIC pn pipeline processing. The user is only advised to check the background level as a function of time, with as main aim the identification of any periods of enhanced low-energy proton flux, see § 4.4.4.

For pn, 13 valid patterns have been defined. As in case of the MOS, pattern 0 events are single pixel events. These comprise most of the valid X-ray events with the most accurate energy resolution. For image analysis patterns 0 to 12 (see figure 17) are the canonical set of valid X-ray events. All higher patterns are not created by single X-ray photons and are due to pattern pileup. For spectral analysis however, only single and double (pattern 0 to 4) should be used, because only these are well calibrated. For the timing mode only singles plus doubles (pattern 0 to 4) should be selected for spectral analysis. Selection of these patterns constitutes the best trade-off between detection efficiency and spectral resolution.

Figure 17: List of valid EPIC-pn patterns (cf. figure 16). Here "." marks a pixel without an event above threshold, "X" is the pixel with the maximum charge ("main pixel"), "x" is the pixel with a non-maximum charge, "m" is the pixel with the minimum charge. These 13 figures refer to the SAS PATTERN codes 0 (singles), 1-4 (doubles), 5-8 (triples) and 9-12 (quadruples), respectively. The RAWX co-ordinate is running rightward and the RAWY co-ordinate running upward.

Users need to beware of the spatial non-uniformity of low energy multi-pixel events, e.g. when defining source and background accumulation regions (§ 4.8.1).