Question ID = 0109
Question ID = 0109
Assigned Member (担当者)2008-09-23 M.NakajimaName and Institute (質問者/機関)Dr. Markowitz (UCSD)Question (質問)Dear Suzaku help desk,I have a question for the HXD Team: I have been re-extracting the HXD PIN and GSO spectra for the relatively bright AGN Cen A, observed 2005 Aug 19-20, using the version 2.0.6.13 pipeline processing events. As shown in the attached two plots, I seem to be getting VERY large data/model residuals associated with the version 2.0 GSO spectrum, but not version pre-1.2-r1 GSO spectrum. (For both GSO spectra, orbits through SAA passage have been excluded. and the source is about ~5-15% of the GSO background.) For example, in the file GSOtest1.ps, the green points are the background-subtracted GSO spectrum published in the first Cen A paper in 2007; that spectrum was made from version pre-1.2-r1 processing; the background was model 'D'. The data/model residuals are within 40%. The red points are the background-subtracted GSO spectrum using the version 2.0 background and version 2.0.6.13 source events file, and the data/model residuals show a very large amount of scatter compared to the version pre-1.2-r1. Reprocessing the version 2.0.6.13 unfiltered GSO source events file according to http://heasarc.gsfc.nasa.gov/docs/suzaku/analysis/hxd_repro.html still yields a spectrum with these +/- 100-200% residuals. (Do the background events files need to be re-processed as well?) In each case, the total spectra (before background subtraction) does not have these very large point-to-point residuals, and neither did the version 1.2 background, leading me to wonder if there's something about the version 2 background I'm not getting. It is not clear to me if I am doing something wrong. Any suggestions as to what could be causing these large residuals? Answer from(回答の文責)Yasushi Fukazawa, Kazuhiro Nakazawa and the HXD-teamAnswer (答え)We, the HXD team, realized this problem only last month.Thank you for pointing this out. As an immediate answer, we suggest you to use the version 1.2 background, together with the version 1.2 data. Following is the detail. ---- As noted in the Suzaku-memo-2008-01, the GSO background is modeled referring to the actual NXB (and on-source data, as well, if the source is not very bright) light curve. In the version 1.2 BGD, we included the "pre-first light observation" of MCG-6-30-15 when modeling the GSO-BGD. In the ver 2.0 BGD case, we did not used it because of unknown reason (maybe the observation is considered too early??). Although we did not realized the difference till now (we regret we did not checked it in detail), your result (and our analysis, as well) shows that the ver1.2 BGD is better to use. Although there are minor differences, the bgd_d of ver 1.2 is very "similar" to the ver2.0 BGD. We are planning to replace ver2.0 BGD after correcting this error, though it will take some time. >From these reasons, meanwhile the ver1.2 BGD could be still useful for your analysis. Status （詳細なステータス）2008-09-23 Accept2008-10-06 Reso. 2008-10-06 Done 
|