Assigned Member (担当者)

Name and Institute (質問者/機関)

Question (質問)

• Question 1
  I want to check the status of the calibration in the observation of NGC 5548 (Obs ID 702042010-702042080) by fitting the spectra of the calibration source at the corners of the CCD chips. I used a Gaussian line to fit the Mn Ka line at 5.895 keV with the data in 5.6-6.2 keV band and found the width, \sigma=12 +3/-5 eV (90% error). But I found the residual at the wings of the Gaussian line (see the attached). If I fitted the Mn Ka line only with the data in 5.75-6.04 keV band, the \sigma is zero and the 90% upper limit is only 11 eV. According to the answer to question ID=0118 in the Suzaku Help Desk, we need not to subtract the background when fitting the spectra of the calibration source. So is the residual at the wings due to the intrinsic non-Gaussian shape of the Mn ka line? And what is the appropriate method to determine the calibration residual of the width? Thank you.
• Question 2
  I find in Section 5.1 of Koyama et al. 2007, PASJ, 59. S23 that 'We note, however, that the spectral resolution plotted in figure 13 is measured at the edge of the detector. The resolution is slightly better at the center of the array, where most targets are imaged, because events detected at the center of the array experience fewer charge transfers.'.
  What is the approximate number of the difference? For example, if the energy resolution at the center is 150 eV (FWHM), what is the energy resolution at the edge? Thanks.

Answer from(回答の文責)

Answer (答え)

• Answer 1
  The residual of your fitting of the XIS data of the calibration source region with the Gaussian model is NOT caused by some intrinsic non-Gaussian shape of Mn Ka line(s) BUT some other effects on energy response matrix of the XIS, for example, due to probability of split events between pixels. The value of energy resolution you got, 13 eV, is within the calibration uncertainties of the XIS, which also includes systematic errors of the pulse-height shape of energy response of the XIS. If you need to know more details, please read the paper; Koyama et al. 2007, PASJ, 59. S23. This is the answer from the XIS team, so please go ahead to analyse your data recognizing on these systematic errors.
  
  Additional Comment
  Here, we (the suzaku helpdesk and the XIS team) did not follow your condition of the analyses of your report. How did you make the plot you attached?
  • which sensor you use on the fitting
    (did you add all the pha files of whole XIS sensors?)
  • which rsp file you use
    (did you use weighted averaged rsp files, if you use sum of XIS pha file?)
  We recommend to check pha and rsp file of each XIS data, separately.
• Question 2
  The position dependence of the energy resolution of the XIS is smaller than the systematic error of its calibration. Thus, the ftool, xissimarfgen, also ignore the position dependence of the energy resolution if you consider the systematic errors.

Status （詳細なステータス）