Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://xmm.vilspa.esa.es/docs/documents/CAL-SRN-0119-1-2.ps.gz Äàòà èçìåíåíèÿ: Wed Jun 26 13:08:35 2002 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 22:13:19 2012 Êîäèðîâêà:

XMM­Newton CCF Release Note
XMM­CCF­REL­119
OM PSF
Bing Chen
June 13, 2002
1 CCF components
Name of CCF VALDATE List of Blocks
changed
CAL VERSION XSCS flag
OM PSF1DRB 0006 2000­01­01T00:00:00 PSF­U No
PSF­B No
PSF­V No
PSF­UVW1 No
PSF­UVM2 No
PSF­UVW2 No
2 Changes
This CCF file provides new measurement of OM PSF. For the U, B and V filters, the PSF is stored
for different count to framerate ratios (CFRR). Since most of the sources in the UV filters are
relatively faint, only a single PSF (so CFRR independent) in UVW1, UVM2, UVW2 is available
now. Moreover, we have no enough data to derive the UVW2 PSF accurately, new UVW2 PSF is
copied from UVM2 PSF as an approximation.
3 Scientific Impact of this Update
The current PSF were generated from Ground Calibration data. This new PSF has been derived
from in­flight data, taking coincidence loss and CFRR into account. This update will improve the
accuracy of the OM photometry, both for UV and optical filters.
1

XMM­Newton CCF Release XMM­CCF­REL­119 Page: 2
Figure 1: Growth curves as a function of radius for V and UVW1 filters. The dashed lines are the
old PSFs, and the symbols are from the new PSF. For OM, one subpixel is about 0.48 arcsec. The
different symbols in the V filter represent the different cfrr groups.
4 Estimated Scientific Quality
The growth curves have been updated significantly using in­flight observations. In figure 1, we show
the growth curves as a function of radiu s for V and UVW1 filters from old CCF (or current CCF)
and the new one. The growth curves for U, B and UVM2, UVW2 are similar to that for V, and
UVW1, respectively.
From Figure 1, we can see that the current PSF from pre­flight observation is very narrow. This
is because that the accuracy of this pre­flight PSF is hampered by the uneven illumination of the
detector and the change in PSF width originating in the CSL collimator.
While this new CCF is stable and sufficiently good, we cannot make a field position dependent
PSF with our data. Therefore, this new PSF is field position independent. Updates are expected
once we have more data and a better understanding of OM large scale sensitivity variations.
5 Test Procedures
This new CCF has been extensively tested using SAS version 5:3:3 at xvsas01.vilspa.esa.es. Several
datasets in Rev 196, 202, 268, 338, including default image mode, Full frame low resolution mode,
have been run through SAS omichain tasks. No error message has been detected.

XMM­Newton CCF Release XMM­CCF­REL­119 Page: 3
Figure 2: The magnitude difference between the new PSF and the old PSF as a function of aperture
radius using the latest SAS 5.3.3
In order to check the influence of the new PSF on the final photometry, in next section, we com­
pared the magnitude derived from the current (om psf1drb 0004.CCF) and new (om psf1drb 0006.CCF)
PSF as a function of the aperture radius.
6 Summary of the test results
For the UV filters, SAS measures counts within a 12 subpixel aperture (e.g. 6 arcsec), then applies
coincidence loss correction and aperture correction to get counts with a radius of 35 subpixel. Here
we show the results for UVW1 filter from an observation in rev. 196. This is a default image
mode. Figure 2 shows the magnitude difference between psf1drb 0006 (new PSF) and psf1drb 0004
(old one) as a function of aperture radius. We can see, for a 12 subpixel aperture, the magnitudes
derived from the new PSF are about 0.085 mag brighter. Because the current UV PSF is much
narrow and reaches 100% with a 5 subpixel aperture, no aperture correction is needed if one uses a
12 subpixel aperture. On the contrary, the new UV PSF is much wider and extends to 35 subpixels,
an aperture correction is needed, which makes the objects brighter. Figure 2 also shows that the
aperture correction term depends on the aperture radius used.
We should point out, due to a bug in SAS v5.3, the UV magnitudes derived from SAS5.3 are
about 0.17 mag fainter. This can be seen in Figure 3, where I compared the magnitude derived with
SAS 5.3.3 and SAS 5.3 using new UV PSF. This bug has been fixed in the latest SAS 5.3.3.

XMM­Newton CCF Release XMM­CCF­REL­119 Page: 4
Figure 3: The magnitude difference in UV filters between SAS 5.3.3 and SAS 5.3 using new UV
PSF.
Figure 4: The magnitude difference in V filter between the new PSF and the old PSF as a function
of aperture radius using the latest SAS 5.3.3

XMM­Newton CCF Release XMM­CCF­REL­119 Page: 5
For the U, B, V filters, the default aperture is 12 subpixels. In the current SAS version, the
aperture radius is independent of the brightness of the source, but it is reduced in the crowded
fields. Here, we present a set of results from ODF 0134562201 in rev. 338. This is a low resolution
full frame observation in V filter. Figure 4 show the magnitude difference between psf1drb 0006
and psf1drb 0004 as a function of aperture radius. For the majority of the stars where the default
aperture (12 subpixels) is applied, no magnitude difference has been found. However, for the stars
in the crowded field where a smaller aperture (! 12 subpixel) is applied the magnitudes from the
old PSF are systematically fainter. The reason is clear. As I have explained above, the current
optical PSF is narrow and reaches 100% at 4 subpixels, and the new PSF is wide and only reaches
100% flux at 12 subpixels (see Figure 1). Therefore, when one uses a small aperture radius ( 4 ! r
! 12 subpixels), no correction is needed for the current PSF, but an aperture correction is needed
for the new PSF, which makes the objects brighter.
In summary, this CCF update has been made correctly and the test results are what we have
expected.
7 Acknowledgements
Thanks to OM team members, especially Igor Antokhin(ULG) and Alice Breeveld (MSSL) for their
inputs.