SUZAKU
Top
Suzaku
Suzaku Flash
Research
Research News
SWG
Tech Description
Presentation
Proposal
Accepted Targets
Schedule
Plan
Log
Process
Archive
Analysis
CALDB
Help Desck
HXD-WAM
Public Data
Bibliography
SWG
Japanese
Top > Suzaku > Analysis > XIS > window

A new recipe for generating NXB background spectra.

2015-04-24 by the XIS team
2015-11-30 by the XIS team

Introduction

In the standard processing of the XIS data, flickering pixels are removed. Flickering pixels are defined as those exceeding a threshold in a given period of time, thus the number of flickering pixels depends on how long the data are integrated. For the NXB database, a very long integration time is used (19 months), whereas individual observations are much shorter. Because of this, the NXB spectra generated using xisnxbgen are lower than actual background spectra. This was not a problem in the early phase of the mission when the fraction of flickering pixels were tiny. However, the fraction keeps increasing, making the discrepancy noticeable recently.
Here, we present a workaround for this problem. A cumulative flickering pixel map ("noisy pixel map"; the map of the pixels that have experienced flickering more than once throughout the mission in the past) is provided in CALDB. This map should be applied for observation files to better match the background spectra generated by xisnxbgen at a sacrifice of an effective area up to 6.5%.

Pros and Cons of this method

  • A significant improvement to reproduce the NXB spectra below 1 keV.
  • The improvement is larger for newer (more recent) data.
  • The efffective area will be decreased up to about 6.5% by applying this method.
  • Users can choose between the conventional method (by default) or the method presented here.
This method is left as an option for users. In general, we recommend this method for weak sources with low surface brightness.

Recipe

Previously, we released a method using a region file. With this method, the decrease in the effective area was not reflected in the ARF files. We thus provide a revised method as follows.
  1. For science observation data,
    1. Set the flag for the but 22 of the pixel quality for an EVENTFILE.
      > pset xisputpixelquality badcolmfile=ae_xi?_npmsci?_20150717.fits
      > xisputpixelquality EVENTFILE.evt
            
    2. Remove noisy pixels from an event file and save it into a new file. This example assumes that events with the pixel quality bit 0--18 are selected. The bit 22 is unselected, thus the resultant file does not include the noisy pixels.
      > ftcopy 'EVENTFILE.evt[EVENT][STATUS=0:524287]' EVENTFILE_NEW.evt
            
    3. Apply to xissimarfgen.
      > pset xissimarfgen badcolmfile=ae_xi?_npmscion_20150717.fits
      > xissimarfgen instrume=XIS? source_mode=SKYFITS source_image={imagefits} pointing=AUTO num_region=1 region_mode=SKYREG regfile1={regionfile} arffile1={arffile} limit_mode=MIXED num_photon=100000 accuracy=0.005 phafile={phafile} detmask={calmask} gtifile={eventfile} attitude={attfile} rmffile={rmffile} estepfile=default
            
  2. For background data by the NXB database,
    1. Set the flag for the but 22 of the pixel quality for a selected NXB event file in the CALDB.
      > pset xisputpixelquality badcolmfile=ae_xi?_npmsci?_20150717.fits
      > xisputpixelquality ae_xis?_nxbsci?_${DATE}.fits
            
    2. Remove noisy pixels from an event file and save it into a new file. This example assumes that events with the pixel quality bit 0--18 are selected. The bit 22 is unselected, thus the resultant file does not include the noisy pixels.
      > ftcopy 'ae_xis?_nxbsci?_${DATE}.fits[EVENT][STATUS=0:524287]' ae_xis?_nxbsci?_${DATE}_rejectnpm.fits
            
    3. Apply to xisnxbgen.
      > pset xisnxbgen nxbevent=ae_xis?_nxbsci?_${DATE}_rejectnpm.fits
            

Demonstration

Below, some comparisons are shown between the NXB spectra before and after applying the method described above.
  • A sample image of before (right) and after (left) removing noisy pixels shown with a square with a red slash.
  • Image
  • A sample spectrum of before (black) and after (red) removing noisy pixels for XIS0 (left), XIS1 (middle), and XIS3 (right).
  • Spectra
  • A sample ARF effective area before (black) and after (red) removing noisy pixels for XIS0 (left), XIS1 (middle), and XIS3 (right). The bottom panels show the ratio between the effective area before and after applying the method.
  • ARF
  • A sample of light curves before (top) and after (bottom) removing noisy pixels for XIS0 (left), XIS1 (middle), and XIS3 (right). Some anomalous bins are removed by applying the noisy pixel removal.
  • Light curve

    ISAS/JAXA Department of High Energy Astrophysics