XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
Recent changes in Epic-related
SAS spectral tasks
Richard Saxton
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
SPECGROUP
Define the useful range of spectral channels Group channels according to statistical criteria Set keywords for spectral fitting programs Currently works on EPIC and RGS spectra (others ?) Enhanced version of GRPPHA
A new SAS task to:
A new SAS task designed to:
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
Setting bad channelsBad ranges pre-defined in XMM_SPECQUAL CCF.
Screenshot
Of
CCF blocks
specgroup spectrumset=myspec.ds setbad=CCF
Or can be set explicitely:
specgroup spectrumset=myspec.ds setbad=“0.0-0.4 10.0-25.0” units=KEV rmfset=epn.rmf
specgroup spectrumset=myspec.ds setbad=“0-10 1100-1200”
MOS, PN
RGS
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
Screenshot
Of
CCF blocks
MOS, PN
RGS
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
Grouping options Manually set the ranges
specgroup spectrumset=myspec.ds ranges=“0-10 11-20 1100-1200” Regular binning
specgroup spectrumset=myspec.ds regbinstart=1 regbinwid=10
Group by counts
specgroup spectrumset=myspec.ds mincounts=25
Group by signal to noise
specgroup spectrumset=myspec.ds minsn=5
Group by ratio to background
specgroup spectrumset=myspec.ds backgndset=mybgd.ds minratio=3 Combination
specgroup spectrumset=myspc.ds minsn=5 ranges=“0:10 1100:1200”
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
End bin options
Set ungrouped pixels bad:
specgroup spectrumset=myspec.ds minsn=5 lastbin=setbad
Add them to the adjacent group
specgroup spectrumset=myspec.ds mincounts=25 lastbin=addtogroup
Set them as an independent group
specgroup spectrumset=myspec.ds minsn=5 lastbin=owngroup
Group from low or high energies
specgroup spectrumset=myspec.ds minsn=4 hightolow=yes
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
Energy resolution Use fraction of resolution defined in CCF
specgroup spectrumset=myspec.ds minsn=5 oversample=3
rmfset=mymos.rmf
This ensures that any one group is
at least as wide as the energy resolution
at the central energy of the group divided
by the oversampling factor.
Energy resolution is stored in the
XMM_SPECQUAL CCF:MOS PN
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
MOS PN
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
Keyword setting
By default the program sets the BACKFILE, ANCRFILE, RESPFILE keywords to either “NONE”or to the values given on the command line.
specgroup spectrumset=myspec.ds arfset=myarf.ds rmfset=myrmf.ds
will set the arf and rmf filenames as given and will set BACKFILE to
NONE.
These are needed for XSPEC 12 to load the file.
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
CAL file issues
We read the energy resolution and bad channel ranges from an XMM CCF for the MOS, PN and RGS instruments. Task should be usable for other instruments, e.g. Chandra. There is a need for a mission-independent calibration file format !
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
• Elliptical parameterisation of the 2-d stacked PSFs. Released in SAS 10.
• Parameters stored in XRTn_XPSF_010.CCF files
- King fn: core radius, slope, ellipticity
- as function of (instrument), energy, off-axis angle
• Phi (azimuthal) dependence = rotation of ellipse
ELLIPTICAL King function
0°45° 90°
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
• Elliptical parameterisation of the 2-d stacked PSFs. Released in SAS 10.
• 1-D PSF ‘ELLBETA’ mode – King profile (r0, alpha), ellipticity – fn(instr, E, theta) + Gaussian (FWHM, norm) – used in ARF generation/spectral fitting for elliptical regions.
• Phi (azimuthal) dependence = rotation of ellipse
• ---------------------------------------------------------------
• eregionanalyse calculates the optimum ellipse
• arfgen recognises the optimum ellipse region and calculates the encircled-energy correction using the ELLBETA model. A non-optimum ellipse uses the ‘HIGH’ accuracy PSF model.
• These tasks packaged up in xmmselect “OGIP spectral products’
Producing optimum ellipse
XMMEPICSOC
Richard SaxtonBackground, Operations & Calibration (BOC) meeting
Madrid, 23/03/10-24/03/10
efluxerefluxer spectrumset=mysrc.ds arfset=mysrc.arf rrmset=mysrc.rmf
backgndset=mybgd.ds fluxedset=mysrc_fluxed.ds