XMM-Newton Science Analysis System


eexpmap (eexpmap-4.12.1) [xmmsas_20230412_1735-21.0.0]

Matching exposure maps and event coordinates

Earlier versions of eexpmap often produce exposure maps with zero exposure at certain image pixels where the event count is nonzero in the input image. This is especially so when the input image is produced from an EPIC event list with randomized coordinates (the default), and for sky images. This usually appears as an offset between the input image and the exposure map, but in no particular direction, and no shifting between the two can make them match exactly. For exposure maps in detector coordinates DETX/Y, the maximum `offset' is $\pm 1$ image pixels (default: $\pm 4$ arcsec). For sky maps, the maximum offset depends in part on the attitude information specific to each observation, but is usually less than $\pm 2$ image pixels.

Bugs related to this problem have mostly been fixed since eexpmap 4.6.1 of xmmsas 9.0.1. eexpmap should now produce exposure maps in DETX/Y that match input images exactly. The single exception is when the event list from which the input image is made contains any PN event with RAWX=28 and DETX=5476, or with RAWY=72 and DETY=9588. The DETX/Y values in these cases actually lie outside the RAWX/Y pixels. Removing these events, reassigning their DETX/Y, or turning off randomization will all solve the problem.

To produce exposure maps in sky coordinates that match input images exactly, users should call eexpmap with the parameter attrebin set to at most $10^{-7}$ radian, i.e., attrebin=0.020626481 (arcsec) or smaller. For long observations, this could prolong the running time of eexpmap substantially. However, exposure maps thus produced will match the input images exactly. Doing the same with older versions of eexpmap will also help to reduce, but will not completely eliminate this offset.