Updated Calibration of Non-univorm CTI
As reported earlier, the column-to-column variations of the charge transfer inefficiencies (or "non-uniform CTI") has the effect of degrading the energy resolution of SIS data. We have incorporated updated values into a new version of the CTI calibration file.
The Original Calibration
The SIS team used the calibration observations of Cas A performed during 1993 - 1995 July to evaluate the degree of energy resolution degradation due to non-uniform CTI. This was then converted into an effective increase fano factor, which was then estimated to be 2.7 x 10-9 s-1 for SIS-0 and 3.3 x 10-9 s-1 for SIS-1. These were further converted into effective line width, and included in the series of CTI calibration files, starting with sisph2pi_290296.fits up to the recently released sisph2pi_130201.fits.
The SIS response generator, sisrmg (version 1.1), reads a CTI calibration file, and broadens the response, interpolating the tabulated values for the epoch of your observation.
The New Calibration
The SIS team has revised the calibration of the non-uniform CTI effect to fit the Cas A calibration observations up to 1999 July, namely with a baseline of 6.5 years rather than the previous 2.5 years. (Strictly speaking, the new values have been derived using data up to 1998 July, but 1999 July data are consistent with these; unfortunately, we could not perform similar observations in 2000 July due to the attitude control problem.)
The revised values are 3.3 x 10-9 s-1 for SIS-0 and 2.9 x 10-9 s-1 for SIS-1. To reduce the potential for future confusion, we have edited in these new values into Dr. Dotani's article in ASCAnews Issue 5 as well as into the Data Reduction Guide
When recent observations are analysed using the old values, this might have noticeable effects, particularly for bright line-rich objects observed in 1-CCD mode. (Note that, for 2-CCD and 4-CCD mode data, the RDD effect results in additional degradation of spectral resolution; this effect is separately taken into account by sisrmg. This would dilute the effect of using an incorrect calibration of non-uniform CTI.)
New CTI File
We have created a new calibration file, reflecting these revised values: sisph2pi_290301.fits.
This corrects for the problem described above, and contains the current best model of the gain calibration.
To generate new response matrices with this calibration,
- Only the rmf file needs to be regenerated using the current version of sisrmg, but specifying the new CTI file.
- It is easiest if someone with a write permission can download the
new CTI file into your FTOOLS refdata/sisdata directory. In this case,
invoke sisrmg as:
sisrmg phtopi=sisph2pi_290301.fits
- If not, copy the contents of the refdata/sisdata directory into
your own directory, as well as the new CTI file. If that directory
is called /home/myname/mysisdata, then invoke sisrmg
as:
sisrmg datadir=/home/myname/mysisdata/ phtopi=sisph2pi_290301.fits
Future Works
As Dr. Dotani's page at ISAS states, the resolution degradation is actually position dependent. The above fix applies to targets observed at the nominal 1-CCD position. A new version of sisrmg that incorporates the position-dependent resolution degradation formula (as well as several other calibration updates) is in preparation.
If you have any questions concerning ASCA, visit our Feedback form.