Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://xmm.vilspa.esa.es/docs/documents/CAL-SRN-0003-1-0.ps.gz Äàòà èçìåíåíèÿ: Wed Jan 3 15:10:22 2001 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 21:29:06 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: òóìàííîñòü àíäðîìåäû

XMM­Newton CCF Release Note
XMM­CCF­REL­3
EPIC ASTROMETRY
D Lumb
October 3, 2000
1 CCF components
Name of CCF VALDATE List of Blocks
changed
CAL VERSION XSCS flag
EMOS1 LINCOORD 0012 2000­01­01T00:00:00 LINCOORD FOV YES
EMOS2 LINCOORD 0012 2000­01­01T00:00:00 LINCOORD FOV YES
EPN LINCOORD 0008 2000­01­01T00:00:00 LINCOORD FOV YES
XMM MISCDATA 0011 2000­01­01T00:00:00 MISCDATA YES
XMM BORESIGHT 0013 2000­01­01T00:00:00 BORESIGHT YES
2 Changes
The CCF files, LINCOORD, contain the data for CCD geometries and MISCDATA includes the focal
length (translating to the mm to arcseconds conversion), while the file BORESIGHT includes the
pointing direction alignments between star tracker and instruments. LINCOORD and BORESIGHT
generally have to be updated together.
The PN camera comprises a monolithic CCD, so that this geometry is well understood, and
confirmed with ground measurements. The focal lengths were determined on­ground, but found
to be 0.1% shorter than predictions, and the cause of discrepancy never satisfactorily explained.
Measurements on several astrometric fields have remnant statistical and systematic errors which are
of the same order.
The current analysis is limited by a number of factors, these include:
ffl the reported star tracker pointings have not been translated systematically into valid ODF
formats, and many analyses have been conducted on the basis of REQUESTED pointings.
1

XMM­Newton CCF Release XMM­CCF­REL­3 Page: 2
ffl Furthermore the detailed valid Attitude History Files have not yet undergone any re­calibration
for guide star locations within the field, which in the ISO programme were found significantly
to improve location accuracies.
ffl The departure from circular symmetry of the Point Spread Function with off­axis angle can
produce a systematic shift in the simple centroid calculation
ffl The CCD focal planes do not match the ideal curved focal surface: The PN focal plane is flat,
while the MOS focal plane has stepped locations of the off­axis CCDs, in a direction parallel
with the optical axis.
ffl The telescope tube itself will out­gas, and may be subject to micro­cracking of its carbon fibre.
Trend analysis will be required to determine if there is a long term shift in boresights.
We expect these files to be updated with improving knowledge of the MOS CCD locations and
systematic trend analysis of boresight calibrations
3 Scientific Impact of this Update
First release
4 Estimated Scientific Quality
At the time of writing the PN camera astrometry files are being updated to reflect the improving
knowledge, and residual location errors of a source centroid may be typically 1 ­ 2 arcseconds.
The MOS camera is subject to an additional uncertainty, in that the 7 CCDs per camera were
individually mounted, and the distance between CCDS and rotations must further be accounted for.
As a consequence, the existing astrometry files are not as well­determined as PN. We suggest the
GO assume position locations to 3 ­ 5 arcsecs in some CCDs.
Due to a limit in the Absolute Measurement Accuracy of the AOCS system, it is expected that
for any single observation there can be a significant difference between actual and reported boresight
of order 3 arcseconds.
The field of view is defined by the filterwheel structure. Potentially small variations in the stop
locations might change the actual field of view as seen in detector co­ordinates. The existing field
of view data used for expsoure maps may have small inaccuracies at the ¸1 pixel level

XMM­Newton CCF Release XMM­CCF­REL­3 Page: 3
5 Expected Updates
Further analysis of several fields is under way, and these will allow a refined set of calibrations files,
as available. Thereafter the limitations will be set by the AOCS sub­system and extensive trend
analysis of its systematic deviations will be required to make a major improvement.
6 Acknowledgements
Thanks to EPIC consortium mebers, especially Mike Denby for inputs