Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/docs/documents/CAL-SRN-0167-1-1.ps.gz
Дата изменения: Thu Jun 24 17:24:22 2004
Дата индексирования: Mon Oct 1 22:26:11 2012
Кодировка:
Поисковые слова: storm

XMM-Newton CCF Release Note
XMM-CCF-REL-167
PSF of the X-ray telescopes
A.M. Read
16 June 2004
1 CCF components
Name of CCF VALDATE List of Blocks
changed
CAL VERSION XSCS ag
XRT1 XPSF 0007.CCF 2000-01-01 KING PARAMS NO
XRT2 XPSF 0007.CCF 2000-01-01 KING PARAMS NO
XRT3 XPSF 0006.CCF 2000-01-01 KING PARAMS NO
2 Changes
New analysis has rened the values stored in the King function parameterisation of the
3 EPIC telescope point spread functions (PSFs), i.e. XRT1, XRT2 & XRT3. They are
stored in the KING PARAMS extension of the CCF, and are tabulated as functions of
ENERGY and THETA (o-axis angle).
This analysis has dealt with purely on-axis sources, and the on-axis King parameters have
been updated accordingly. The o-axis behaviour previously observed has been used to
extrapolate the new on-axis parameters to new projected o-axis values.
3 Scientic Impact of this Update
The PSF is described by a King function whose parameters, core radius r 0 and index ,
are themselves functions of energy and o-axis angle:
PSF King (r) = A
(1 + ( r
r0
) 2 )
Earlier work [1, 2, 3] used many bright point sources both on and o axis to determine the
energy dependent PSF. This resulted in a linear dependency of r 0 and with energy and
1

o-axis angle. It is shown here that this linear dependency is not valid the dependencies
of r 0 and are seen to be atter (almost constant) with energy (at least out to
8 10 keV).
3.1 Scientic Impact: Analysis and Results
The present analysis has made use of a small number of newer datasets that have become
available, involving very long and clean observations (with few, if any, periods of high-
background aring) of very bright on-axis point sources. The data have all been taken
in small window mode, such that these very bright sources are not signicantly piled-up.
Specically, the most useful targets that have been used are listed (together with their
revolution numbers) as follows:
MCG-06-30-15 (Revs. 301, 302, 303 [each 100 ks])
Ark120 (Rev. 679, [ 100 ks])
3C273 (Revs. 94, 95, 96, 277, 370, 373, 472, 554, 563 [various exposures, just MOS])
Two threads of analysis were followed: One involved the forming of narrow-energy-band
images from the brightest of these sources, and tting the surface brightness radial proles
obtained from these images with a King function to obtain r 0 and as a function of energy.
A second analysis thread involved the extraction of spectra from narrow annuli around
each point source, and once ARF les had been generated (this involving the actual form
of the PSF), the spectra were tted with standard spectral models, to see how (if at
all) the spectral parameters obtained varied with extraction radius (for a point source
of course, they should not vary at all). This whole process was repeated for several sets
of PSF parameters (including those obtained from the surface brightness radial prole
tting described above).
In forming the narrow-band images, the MCG-06-30-15 data were predominately used.
The datasets were rst cleaned of times of high background, and FLAG=0 single-event
images were created in X/Y coordinates (rather than RAWX/RAWY or DETX/DETY
coordinates), to take into account eects of any small attitude shifts. Dierent energy
binnings were used to get better statistics in the dierent energy bands for the radial
tting (hence the CCF parameters are now mapped to dierent energy values than in
previous versions).
Before constructing the radial surface brightness proles, the centre of the emission was
found for each image via Gaussian tting. The very small (and probably non-signicant)
shifts in Gaussian centre with energy were taken into account in constructing the radial
proles, which were then tted with a King function. Two examples of surface brightness
radial proles plus tted King proles are shown in Fig.1.
This resulted in r 0 and values for 10 dierent energies, for the three dierent instruments
and for the three MCG-06-30-15 observations. Exposure-weighted mean values (over the
2

Figure 1: Surface brightness radial proles (crosses) plus tted King proles (lines) for two examples:
(left) MCG-06-30-15 Rev. 303 pn at 6 keV and (right) MCG-06-30-15 Rev. 302 MOS2 at 0.475 keV
three dierent observations) for r 0 and were then calculated for each instrument and
energy band. A separate analysis involving the 'stacking' of the separate relevant images
from the dierent observations on top of one another, prior to the radial prole tting,
produced almost identical results to the values obtained via calculating the exposure-
weighted mean values.
3.2 Scientic Impact: Application
The resultant dependencies of r 0 and are seen to be atter (almost constant) with
energy, at least up to 8 10 keV (where the r 0 E and E relationships turn over)
than in the previous parameterization of the PSFs. This is shown in Fig. 2.
The new PSFs were used in the analysis of spectra extracted from narrow concentric
annuli around a number of bright point sources, as described above. Fig.3 shows how
the tted normalization and power-law index vary as a function of extraction 'radius' (a
circle of 0 5 00 , then annuli of 5 10 00 , 10 15 00 , 15 20 00 etc.) for the current CCF PSFs and
the new CCF PSFs described here, for the MCG-06-30-15 Rev. 302 data. A point source,
of course, should show no variation in tted spectral parameter whether the spectrum
is extracted from the very centre of the distribution or from the wings, but usage of
the current PSFs result in a very wide range in spectral parameters for dierent radii.
Usage of the new PSFs gives rise to a very signicantly improved situation, with the tted
normalization and power-law index remaining constant and ' at' with radius. Fig.4 shows
3

Figure 2: (top) r 0
Energy and (bottom) Energy dependencies for the new MOS1, MOS2 and pn
on-axis PSFs.
4

Figure 3: Plots showing how the tted normalization (top) and power-law index (bottom) vary as a
function of extraction 'radius' (see text), using the current CCF PSFs (left) and the new CCF PSFs
(right) for the MCG-06-30-15 Rev. 302 data.
the equivalent plots, this time using the Ark120 Rev. 679 data. Again, a very signicant
improvement is seen with the new PSFs.
4 Estimated Scientic Quality
A major problem with the previous parameterisation was its inability to produce con-
sistent spectral ts for annular extraction regions such as are used for the analysis of
piled-up sources. To quantify the improvement with the new PSFs over the old PSFs,
MCG-6-30-15 (from Rev 302) has been extracted from annuli of 5-40 00 , 10-50 00 and 15-60 0
0 and ts compared to those of a circular extraction (0-30 00 ). This has been performed using
the new and the old PSFs, and the results are presented in Fig.5. Whereas usage of the
old PSFs results in a per instrument normalization variation of up to 40%, and changes
in the tted spectra slope of 0.2, the new PSFs give rise to normalization variations of
nearer 5% and a spectral slope change of at most 0.03.
As yet, no sources bright enough for this type of analysis to be performed o-axis have
been observed. Indeed, as very long observations in small window mode are necessary,
and as small window mode can only be accessed close to on-axis, then the situation may
remain like this. As such, the general o-axis results of previous work [1, 2, 3] have been
used to transform the new on-axis parameters presented here to projected o-axis values.
5

Figure 4: Plots showing how the tted normalization (top) and power-law index (bottom) vary as a
function of extraction 'radius' (see text), using the current CCF PSFs (left) and the new CCF PSFs
(right) for the Ark120 Rev 679. data.
5 Expected Updates
None are foreseen as regards revising the King parameterization of the PSFs.
6 Test procedures and Results
The changes introduced here directly aect the encircled energy correction which is applied
by the SAS task arfgen. The analysis detailed in sections 3 and 4 has used arfgen version
1.65.11 (SAS 6.0) in conjunction with the new PSF CCFs.
References
[1] S. Ghizzardi, "In- ight calibration of the on-axis and near o-axis PSF for the Mos-1
and Mos-2 cameras", EPIC-MCT-TN-011.
[2] S. Ghizzardi, "In- ight calibration of the PSF for the pn camera", EPIC-MCT-TN-012
[3] R.D. Saxton, "PSF of the X-ray telescopes", XMM-CCF-REL-116
6

Figure 5: Plots showing how the tted normalization (top) and power-law index (bottom) vary as a
function of extraction region (left to right: 0-30" circle, 5-40" annulus, 10-50" annulus, 15-60" annulus),
using the current CCF PSFs (left) and the new CCF PSFs (right) for the MCG-06-30-15 Rev.302 data.
7