2008-12-08 Y.Nakagawa
2008-12-15 Y.Terada
Mr. Yuan Liu (Tsinghua University)
- Question(1):
Q1: I want to extract the spectra of the calibration sources on the corners of the XIS CCDs.To fit the edges of the CCDs well, I use a polygon in DS9 as the source region instead of a circle.Then I use the script 'xisresp' (http://suzaku.gsfc.nasa.gov/docs/suzaku/analysis/xisresp) to produce .rmf and .arf files. However, I find the values of the effective area in the .arf file are all zero!I also find even using the circle region, if the circle is near the edge of the CCDs, the values of the effective area in the .arf file will be zero!If the circle is far from the corners enough (but with fewer photons of the calibration sources), the produced .arf file will be correct.I am sure both the polygon and the circle are well within the edge of the CCDs.Should I modify the script 'xisresp' in this case?Q2: Should I deal with the calibration source as a point source or an extended source?Thank you!
- Question(2):
I want to analyze the emission lines in the spectra of central AGN.
So the peak energy of Mn K\alpha line of the calibration source can
let me know the exact status of the calibration in this observation.
As you said, the arf file is not necessary. So after generating the
spectra and rmf files, I just need to load them into Xspec and fit
the Mn K\alpha line to determine the peak energy ?
- Question(3):
But how accurate is the approx central energy? For example, if I fit the
line with a Gaussian line in Xspec and obtain E=5.893+/-0.005 keV. What
is the systematic error due to the approximation?
- Question(4):
Is it necessary to subtract background (obtained from nearby region)
from the spectra of the calibration source?
Suzaku Help Desk
- Answer(1):
A1: No. If you want to determine central energies of spectra of the
calibration sources, you do not need the ancillary response files
(the ARF files). This is because the ARF files *cannot* be defined
by definition for the calibration sources which photons do not pass
through the X-Ray Telescope (XRT).
A2: Again, you do not need the ARF files for analyses of the calibaration
sources. If you want to analyze objects behind the calibration sources,
the ARF files depend on the objects.
- Answer(2):
Yes. You can only determine the approx central energy of the
calibration source by loading the spectrum file and the RMF file
into XSPEC. This is because a relation between PI (channel) and
energies (keV) is described only in the 3rd extension (EBOUNDS)
of the RMF files.
For your reference, a relation between ENERGY and MATRIX is
described in the 2nd extension (MATRIX) of the RMF files,
while a relation between ENERGY and SPECRESP is described in
the 2nd extension (SPECRESP) of the ARF files.
- Answer(3):
In loading the spectrum files and the RMF files into XSPEC,
there is no systematic error for your analyses to determine
the central energy of the calibation sources. You just have
statistical error. In the energy calibration by the XIS team,
we think that your result E = 5.893 keV has ~2 eV accuracy.
- Answer(4):
No. You do not need to subract the background in this analysis.
2008-12-08 Accepted
2008-12-11 Answer(1)
2008-12-12 Answer(2)
2008-12-16 Answer(3,4)
2008-12-16 Close
|
Question ID = 0118