Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/docs/documents/CAL-SRN-0192-1-1.ps.gz Дата изменения: Mon Aug 22 17:35:02 2005 Дата индексирования: Mon Oct 1 22:36:16 2012 Кодировка: Поисковые слова: m 8

XMM-Newton CCF Release Note
XMM-CCF-REL-192
EPIC MOS response
S.F. Sembay, R.D. Saxton
19 August 2005
1 CCF components
Name of CCF VALDATE EVALDATE Blocks changed XSCS ag
EMOS1 REDIST 0057.CCF 2001-11-25 2002-05-01 CCD REDISTRIBUTION-n NO
EMOS1 REDIST 0058.CCF 2002-05-01 2002-11-07 CCD REDISTRIBUTION-n NO
EMOS1 REDIST 0059.CCF 2002-11-07 2004-01-01 CCD REDISTRIBUTION-n NO
EMOS1 REDIST 0060.CCF 2004-01-01 - CCD REDISTRIBUTION-n NO
EMOS2 REDIST 0057.CCF 2001-11-25 2002-05-01 CCD REDISTRIBUTION-n NO
EMOS2 REDIST 0058.CCF 2002-05-01 2002-11-07 CCD REDISTRIBUTION-n NO
EMOS2 REDIST 0059.CCF 2002-11-07 2004-01-01 CCD REDISTRIBUTION-n NO
EMOS2 REDIST 0060.CCF 2004-01-01 - CCD REDISTRIBUTION-n NO
2 Changes
A small patch has developed at the boresight of each MOS camera where the spectral
resolution of the detectors at low-energies (< 0:8 keV) is degraded [1]. This leads to the
need to quantify the redistribution function (RF) of each camera in three dierent spatial
regions.
 A core region centred on a point between the two boresights where EPIC-pn and
RGS are prime.
 A wings region, consisting of an annulus about the core
 An outer region, consisting of the CCD-1 area which is outside the core and wings
regions.
The redistribution function for the outer CCDs is unaected by this change and has been
set equal to that of the outer region of CCD-1.
Physically, it is believed that the incident X-rays, possibly in conjunction with soft pro-
tons, have damaged the CCDs in such a way that an increasingly large amount of charge
is lost to the surface layer as the mission progresses. At the boresight, where the X-ray
1

Table 1: Regions in RAW coordinates on CCD-1
Camera Core Wings Outer
MOS-1 Circle(312,293,14) Annulus(312,293,14,36) !Circle(312,293,36)
MOS-2 Circle(301,302,14) Annulus(301,302,14,36) !Circle(301,302,36)
doseage has been the highest, this damage has led to a broadening of the RF which is
particularly noticeable at low energies and for later revolutions. This release covers rev-
olution 360 and beyond when the eect is already important. The smaller eects seen
prior to this revolution are in the process of being measured and a further release of CCFs
is expected for earlier epochs.
The spatial regions (Tab. 1) have currently been set to be the same for all epochs which
is something that will need to be reviewed as the mission progresses. It is expected that
the aected regions will gradually expand as greater numbers of X-rays are accumulated.
SAS software and the CCF structure have been designed to cope with changes to the
number and denition of these regions.
An indication of the extent of the time dependence of the core, wings and outer RFs is
show in Figure 1. Here, the reponse at 0.5 keV can clearly be seen to change more within
the patch region.
The time span from 07/11/2002 has been divided into two periods in this release: until
the end of 2003 and from 2004 onwards.
3 Scientic Impact of this Update
It has long been noted that spectral ts below 1 keV show discrepancies between the
EPIC cameras with pn returning more ux than the MOS cameras, sometimes by up to
20% [2]. In addition there is strong evidence that the discrepancy increases with time.
The identication of a time-dependent change in the MOS redistribution parameters for
on-axis sources at low energies improves the agreement between the cameras considerably.
The residual errors are now seen to be around 5{10% for most sources.
The 8 CCFs in this release have been tested extensively in the validation exercise un-
dertaken for SAS 6.5 [3]. In general the results show an improvement in the agreement
between the pn and MOS cameras for post-revolution-360 observations. As an example,
in Figure 2 the residuals of a galactic absorption plus two power-law model t to the pn,
MOS-1 and MOS-2 spectra of a revolution 1023 observation of the Radio-Loud Quasar
3C273 are shown. There is a clear improvement in the agreement between the instruments
below  0:5keV . The reduced  2 improves from 4551/2744 to 3831/2744 with the new
redistribution function.
2

Figure 1: A comparison of the MOS-1 response function for three dierent epochs (revs 1, 360 and 744)
at 0.5 keV. Top panel: core of the patch, Centre: wings of the patch, Bottom: outside the patch
3

Figure 2: An absorbed two-power-law t to a revolution 1023 observation of 3C273. The upper plot
gives the residuals using RMFs generated by SAS 6.1 using the CCFS EMOS1 REDIST 0051.CCF and
EMOS2 REDIST 0051.CCF. The lower plot shows the same t but using responses generated by rmfgen
using EMOS1 REDIST 0060.CCF and EMOS2 REDIST 0060.CCF.
4

4 Estimated Scientic Quality
This update improves the agreement between the EPIC cameras to 5{10% at low ener-
gies for most sources (see [3] for a summary).
5 Expected Updates
Calibration is continuing for data prior to revolution 360. It is expected that CCFs for
these epochs will be released in due course.
6 Test procedures and results
The new CCF les were used to produce redistribution matrices using rmfgen for on patch
and o-patch regions and several dierent epochs. These were compared with canned
matrices produced by the MOS instrument team and were seen to be nearly identical in
all cases.
References
[1] Kirsch, M. et al, 2005, "SPIE 5898-29: Health and cleanliness of the XMM-Newton
science payload since launch", XMM-SOC-CAL-TN-0062
[2] Altieri, B. et al., 2005, "Status of XMM-Newton instruments cross-calibration with
SAS 6.1", XMM-SOC-CAL-TN-0052
[3] Guinazzi, M. et al., 2005, "XMM-Newton Science Analysis System (version 6.5) EPIC
Science Validation Report", XMM-SOC-USR-TN-0009 issue 2.0.
5