Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/docs/documents/CAL-SRN-0264-1-2.ps.gz Дата изменения: Fri May 7 16:48:17 2010 Дата индексирования: Mon Oct 1 22:58:30 2012 Кодировка: Поисковые слова: http astrokuban.info astrokuban

XMM­Newton CCF Release Note
XMM­CCF­REL­264
2­D PSF Gaussian parameterization
R. Saxton
May 6, 2010
1 CCF components
Name of CCF VALDATE List of Blocks changed Change in CAL HB
XRT1 XPSF 0012.CCF 2000­01­01 ELLBETA PARAMS YES
XRT2 XPSF 0012.CCF 2000­01­01 ELLBETA PARAMS YES
XRT3 XPSF 0012.CCF 2000­01­01 ELLBETA PARAMS YES
2 Change
The elliptical beta, 2­D PSF model (ELLBETA) consists of an elliptical King function plus a central
narrow Gaussian to model the core plus spoke structures (see CAL­SRN­0263 for a fuller description).
In some cases no Gaussian component is necessary. In the previous release, the default values of
the Gaussian full width half maximum (FWHM) and normalization for these cases were set to zero.
This has been found to give problems in software which doesn't handle the zeros correctly. To avoid
these problems, the FWHM values for the Gaussian is set to a positive, non­zero value for all the
data points.
. For EPIC­pn the FWHM is set to a nominal value of 0.1.
. For the EPIC MOS 1/2, the values for energies at 8 keV and above, which were previously
0.0, have been set to the FWHM at 6 keV for that particular o#­axis angle. i.e. the 8 keV
Gaussian FWHM value for theta=0 has been set to the FWHM value at 6 keV and theta=0.
This is to maintain a constant FWHM for energies between 6 and 8 keV where the Gaussian
component is linearly interpolated and hence has a non­zero normalisation.
1

XMM­Newton CCF Release XMM­CCF­REL­264 Page: 2
3 Scientific impact of this update
This solves a technical problem, where parts of the SAS software returned NAN values in the PSF
image.
4 Estimated scientific quality
5 Test procedure and results
The new CCFs have been used within the calview and psfgen tasks to generate PSF images for
each of the instruments at a range of energies and o#­axis angles. In all cases the image is seen to
contain real, positive numbers.
The e#ect of varying the default value used for the Gaussian FWHM from 0.1 to 100.0 has been
checked for the EPIC­pn. It was found to make no di#erence to the final image, which is as expected
given that the normalisation is 0.0 in these cases.