Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://xmm.vilspa.esa.es/docs/documents/CAL-SRN-0268-1-1.ps.gz Äàòà èçìåíåíèÿ: Thu Jul 29 20:22:00 2010 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 22:58:49 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: þæíàÿ àòëàíòè÷åñêàÿ àíîìàëèÿ

XMM­Newton CCF Release Page: 1
XMM­Newton CCF Release Note
XMM­CCF­REL­267
EPIC MOS response
R. D. Saxton, M. Guainazzi
29 July 2010
1 CCF components
Name of CCF VALDATE EVALDATE Blocks changed XSCS flag
EMOS1 REDIST 0075.CCF 2004­01­01 2020­01­01 PRIMARY NO
EMOS2 REDIST 0075.CCF 2004­01­01 2020­01­01 PRIMARY NO
2 Changes
In the previous set of EPIC MOS redistribution files, released in December 2005, the end
validity date for the last epoch (EMOS1/2 REDIST 0074.CCF) was set to the beginning of
2010. This is clearly incompatible with observations made during or after 2010. Analysis
of these observations with the current CCF has incorrectly used a much earlier version of
the redistribution files (EMOS1/2 REDIST 0042.CCF; issued in August 2004) which contain
an outdated parametrisation of the MOS photon redistribution.
In this release, the end date (keyword EVALDATE) for the final epoch has been increased
to the beginning of the year 2020 for both MOS­1 and MOS­2. This will ensure that new
observations use the latest epoch CCF element for the expected lifetime of the mission.
3 Scientific Impact of this Update
Only observations performed since the 1st of January 2010 are a#ected.
Examples of the impact yielded by the aforementioned error are shown in Fig. 1, where
we compare the quality of the spectral fitting on two calibration observations performed in
2010: the thermal SuperNova Remnant (SNR) 1E0102­72 (Obs.#0412981001; Rev.#1898)
and the blazar PKS2155­304 (Obs.#0411780501; Rev.#1902). The spectra of the 1E0102­
72 were reduced and analysed following the prescriptions in Plucinsky et al. (2008). The
source spectra of PKS2155­304 were reduced using standard procedures, and extracted

XMM­Newton CCF Release Page: 2
0.1
1
10
Cts
s
-1
keV
-1
Black=MOS1_REDIST_0075; Red=MOS1_REDIST_0074
1E0102-72 - Obs.#0412981001 - Rev.#1898
1
0.5 2
0.8
0.9
1
1.1
1.2
Data/model
Energy (keV)
0.1
1
10
Cts
s
-1
keV
-1
Black=MOS2_REDIST_0075; Red=MOS2_REDIST_0074
1E0102-72 - Obs.#0412981001 - Rev.#1898
1
0.5 2
0.8
0.9
1
1.1
1.2
Data/model
Energy (keV)
10
-3
0.01
0.1
1
Cts
s
-1
keV
-1
Black=MOS1_REDIST_0075; Red=MOS1_REDIST_0074
PKS2155-304 - Obs.#0411780501 - Rev.#1902
1
0.5 2 5
0.8
0.9
1
1.1
1.2
Data/model
Energy (keV)
10 -3
0.01
0.1
1
Cts
s
-1
keV
-1
Black=MOS2_REDIST_0075; Red=MOS2_REDIST_0074
PKS2155-304 - Obs.#0411780501 - Rev.#1902
1
0.5 2 5
0.8
0.9
1
1.1
1.2
Data/model
Energy (keV)
Figure 1: Spectra (upper panels) and residuals (lower panels) in units of data/model ratio when the same
model with the same parameters is applied to the MOS1 (left) and MOS2 (right) spectra of 1E0102­72
(top) and PKS2155­304 (bottom).

XMM­Newton CCF Release Page: 3
from a circular region of 40'' around the X­ray centroid, after excising a 1.5'' PSF core due
to marginal residual level of pile­up. background spectra were extracted from source­free
regions on nearby chips. The PKS2155­304 spectra were analysed with a photoelectrically
absorbed broken power­model.
In each panel of Fig. 1 we show (twice) the same spectrum (upper panels) and the resid­
uals against the same spectral model with the same parameters but when redistribution
matrices generated with di#erent MOS redistribution CCF constituents are used. In
all plots the black curves represent spectra and residuals obtained when the updated
EMOS1/2 REDIST 0075.CCF is used; the red curves represent the residuals obtained when
EMOS1/2 REDIST 0042.CCF is used, i.e. the redistribution CCF constituent wrongly
picked­up by cifbuild instead of EMOS1/2 REDIST 0074.CCF. The di#erence is signifi­
cant, due to the out­of­date calibration enshrined in EMOS1/2 REDIST 0042.CCF. In terms
of data/model ratio against the same spectral model with the same parameters, the rela­
tive di#erence is #5% above #0.5 keV, and can be as high as 10­20% at lower energy.
4 Estimated Scientific Quality
This update is meant to ensure that nominal spectral quality calibration accuracy is
obtained for MOS spectra of sources after January 1, 2010.
Users are warmly recommended to updated their CCF repository with EMOS1/2 REDIST 0075.CCF,
regenerate their Calibration Index Files (CIF) via cifbuild, and then the MOS redistri­
bution matrix via rmfgen.
This error does not a#ect the ``canned'' MOS redistribution matrices available from the
XMM­Newton SOC calibration portal:
http://xmm2.esac.esa.int/external/xmm sw cal/calib/epic files.shtml.
5 Test procedures and results
CIFs were generated via cifbuild using EMOS1/2 REDIST 0075.CCF for several XMM­
Newton observations performed after January 1, 2010. In all cases, the CIF points to the
latest MOS redistribution CCF constituents, as expected.
6 References
Plucinsky P., et al., 2008, SPIE, 7011, 68