[Date Prev][Date Next][Thread Prev][Thread Next][Search] [Main Index] [Thread Index] [HEASARC Mailing List Archives]

Notice on GIS-S2 gain correction

 ------------ Notice on GIS-S2 gain correction.---------------------------
 	     Aya Kubota, Makoto Tashiro, Kazuo Makishima (Univ. of Tokyo), 
 	     Ken Ebisawa (ASCA GOF/ GSFC/NASA)  and the GIS team

The GIS team warns that the GO data observed in December 1999 and
January 2000 may be subject to up to 0.8 % overcorrection of PI values 
in the distributed event files. 
The GIS data obtained in December 1999 and January 2000 and processed 
using "gis_temp2gain.fits (v4.3.18)" have been found to have systematically
higher gain (up to about 0.8%) than usual. This occurred because the GIS-2 
temperature strayed upward from nominal values during December 1999, due to 
particular spacecraft conditions.  This rather large temperature offset, 
in turn, caused unusually large errors when the GIS-2 gain was corrected
through extrapolation, assuming the gain versus temperature relationship
determined using the data till October 1999. 

The new version of gain correction table -- "gis_temp2gain.fits (v4.3.20)" --
updates the gain correction factor based on the data up to December 5, 1999.
We have confirmed that this new version (v4.3.20) corrects the December 1999 
data accurately.  Therefore, we recommend that guest observers with December
1999 and January 2000 data from GSFC re-process their data with the new 
"gis_temp2gain.fits (v4.3.20)" file, if they need energy scales more accurate
than 1 %.  

The gis_temp2gain.fits files is found in the "aux" directory in the data
distribution package, and the GFILENUM keyword in the primary header tells
the version number, which the Guest Observers are encouraged to check.  The 
"ad*_hdr_page.html" file in the "aux" directory also tells the gain file
versions (see the web page indicated below).

For the method to re-determine the GIS gain values of event files, as well as 
for more in-depth information on the GIS gain in general, please refer to