Current SIS Calibration Files
This bulletin was originally released via e-mail exploders on 1994 Nov 9; WWW version 1995 Apr 13.
To coincide with the release of FTOOLS version 3.2, the ASCA GOF has released new sets of:
- SIS response matrices
- SIS blank sky background event files.
These new software and calibration files enable ASCA users to analyze SIS data in PI (Pulse Invariant) rather than PHA space. The chief advantage of using PI is that events from different chips (but same SIS camera) can be combined without compromising the calibration.
This note provides a brief summary of current calibration status and data analysis procedure.
A. Scheduling Requirements
Increasing CTI (Charge Transfer Inefficiency) is a manifestation of radiation damage which became noticeable by early 1994 in the form of a secular gain change. To calibrate this effect, the SIS team produced an initial calibration file (sisph2pi.fits in the refdata area of FTOOLS v3.2). In addition, observations of Cas A were obtained to improve the calibration. The analysis of these data is in an advanced stage.
The same Cas A data also show a discrepancy in energy scale between S0C1 (SIS0-chip1) and S1C3 (SIS1-chip3). In view of both this (also reported from PV phase SNR and Cluster observations) and the secular gain variation, the SIS team is refining the absolute gain calibration of all eight CCD chips. At the moment, there may be a 0.5-1 per cent systematic uncertainty in absolute gain calibration.
B. What is SIS PI?
The FTOOL SISPI (in FTOOLS 3.2) will fill the PI column in your SIS Faint or Bright event files. If you use the default settings:
- The calibration file sisph2pi.fits is used (transparently).
- The PI channels are defined to have 3.65 eV/bin. Bright mode event files and most pha files are rebinned. In such cases, the PI bins are integer multiples of 3.65 eV.
- Differences due to particular chip or position on the chip are taken into account.
- The secular gain drift is corrected.
PI is currently not available for SIS Fast mode data because the PHA-to PI conversion is position-dependent (Fast mode data do not contain positional information).
C. PI columns in your data files
All event files in the public archive will have the PI columns filled in (except Fast mode data). For proprietary guest-observer data, the current processing script does not fill the PI columns, but this will change shortly. In the meantime, GOs should run SISPI themselves. Note that the PI channels of event files may be safely repopulated through multiple runs of SISPI: the original event data are not affected.
D. Spectral extraction procedures
- Run SISPI on the event file(s), if necessary.
- Read the event file into XSELECT. Then type "set phaname PI" in order to ensure that the PI column is used for spectral extractions (the current version of XSELECT uses the PHA column by default).
- Use filters as appropriate. Extract and save a spectrum.
- Extract a background spectrum - either from a source-free region of
the SIS, or from the new blank sky files which are available via
anonymous FTP at heasarc.gsfc.nasa.gov in the directory:
caldb/data/asca/sis/bcf/bgd/94novThese new blank sky background files are essentially unchanged from the previous release in April, except for having their PI columns filled. After reading in a background file into XSELECT, don't forget to "set phaname PI".
E. SIS response generators
After creating source and background PI files, the next step is to obtain the appropriate response. As for PHA files, the spectral response is divided into two parts, the RMF (redistribution matrix file) and the ARF (ancillary response file, i.e., the effective area). When both are combined into a single file, it usually has the extension .rsp.
In general, SIS .rmf files can be generated using the script sisrsp. This provides a friendly interface to the underlying FTOOL, sisrmg. FTOOLS version 3.2 contains sisrmg version 0.8, which is different from the previous version 0.6 as follows:
- High-PH tail model for partial charge collected events is now included. (This has the harmless side-effect of producing the "log: DOMAIN error" message when run on Suns, but the output matrices are still valid).
- Now uses the ISAS QE (Quantum Efficiency) determination for individual chips, based on observations in Dec 93 of 3C 273.
- PI matrices can now be generated.
- The secular gain change due to CTI can now be corrected (using
the same sisph2pi.fits calibration file as SISPI). This means that
you can combine either:
- a PI spectrum with a PI response, or
- a PHA spectrum with a PHA matrix generated with the correct gain for the epoch of observation. When data from only one chip are used, this is equivalent - for data analysis purposes - to using a PI file with a PI response.
For detailed instructions, type "sisrsp" without arguments.
In addition to the secular gain change, there have been smaller changes in the SIS resolution. The Calibration of this effect is underway but is not incorporated in the v0.8 matrices.
For ARF generation, use the FTOOL ASCAARF which takes as input the RMF (PI or PHA) and the spectral file (PI or PHA). ASCAARF will then create the .arf file which has the effective area curve calculated for your extraction region.
F. Ready-made response matrices
For your convenience, we are providing a 'standard set' of SIS matrices in directory caldb/data/asca/sis/cpf/94nov9 via anonymous FTP at legacy.
For all chips:
s0c0g0234p40e0_1024v0_8h.rmf&s0c0g0234p40e0_1024v0_8i.rmf: Grade 0234 Bright2 mode 1024-channel PHA (h) and PI (i) matrices.s0c0g0234p40e0_512v0_8h.rmf&s0c0g0234p40e0_512v0_8i.rmf: Grade 0234 Bright2 mode 512-channel PHA (h) and PI (i) matricess0c0g0234p40e1_512v0_8h.rmf&s0c0g0234p40e1_512v0_8i.rmf: Grade 0234 Bright mode 512-channel PHA (h) and PI (i) matrices
For S0C1 and S1C3:
s0c1g02p40e1_512v0_8h.rmf&s0c1g02p40e1_512v0_8i.rmf: Fast mode 512-channel PHA (h) and PI (i) matrices, Fast mode grade 0.s0c1g0234p40e1_512_1av0_8i.rsp: Typical rsp file for a point source observation (rmf & arf combined)
Koji Mukai & the ASCA GOF with Geoffrey B. Crew & the SIS team
If you have any questions concerning ASCA, visit our Feedback form.