2007-03-26 A.Bamba
Lindsay Rocks (University of Wisconsin)
I am Dan McCammon's grad student and I have been working with some of the
Suzaku XIS data for my thesis. I have some recent fits to XIS data done
in XSPEC where the folded model resolution seems worse than than the
actual resolution of the XIS data. I think that the RMF file may be
underestimating the resolution of XIS. I am using an RMF made with
xisrmfgen version 2006-11-26. I made the RMF with a spectrum XSELECT
created with rev 0.7 data.
I have attached a postscript plot of the spectrum, showing the model with
poor resolution as well as the spectrum I am fitting. I will also copy
the header from the matrix extension of the rmf.
Could you please let me know if you know of a way to fix this resolution
problem? I am very anxious to find a solution, please let me know if I
can send you any more information or help in some way.
XIS team
We checked the energy resolution of XIS1 in the low energy band.
Details of the check results are given below.
In conclusion, the energy resolution in
the current RMF is reasonably good, and there is no indication
of large discrepancy in the resolution as you suggested.
On the other hand, one of the line-like features in your
spectral file has unusually narrow width. We urge you to check
the data carefully to find out the origin of the feature.
We used the calibration target, SNR E0102, to check the energy
resolution of current RMF, as the source was observed frequently.
We show in figure 1 an example of the fitting result to E0102 data
observed on 2006 Feb. 02. We used the spectral model obtained
with the Chandra and XMM observations
(eg. A&A, 365, L231; ApJ, 605, 230)
and used the RMF created with "xisrmfgen".
No extra line widths are introduced in the model.
Although the fit residuals show some structures, they are rather
small and the line profiles are reproduced reasonably well.
This means that the energy resolution of the RMF matches the
observed line widths well.
We also fitted your pha file with a simple power-law plus
gaussians model. The result is shown in figure 2.
Two prominent lines at 0.65 keV and 0.8 keV are reproduced well.
We conclude that the energy resolution of the RMF and the observed
line profiles are consistent at least judging from these two lines.
On the other hand, the line feature at 0.73 keV is unusually
narrow, much narrower than the lines at 0.65 keV and 0.8 keV.
Clearly, the line profile cannot be reproduced with the current RMF,
and may be difficult to consider to be intrinsic to the source.
We urge you to check whether the line is really intrinsic to
the source. We believe it is useful to check the image and
the light curve created in a narrow energy band around 0.73 keV
for this purpose. If the feature is considered to be intrinsic
to the source, please let us know. We really appreciate such
information to improve the calibration of XIS, and are willing to
work with you to solve the problem.
2007-03-26 Accept
2007-03-26 Forwarded to XIS team
2007-04-13 Done
|
