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

Two items regarding the GIS

*  New GIS blank sky and night earth background event files available  *

ASCA GOF has released new GIS blank sky and night earth background event files.
The blank sky event files are accumulated from numerous high Galactic latitude
observations through 1993 June to 1995 December, achieving total 2.1 Msec
exposure time.  Point sources brighter than ~2e-13 erg/s/cm2 (2-10 keV) have
been removed.   A new ftool, 'mkgisbgd' can be used to correct exposure,
taking into account the effect of removing the point sources.  Mkgisbgd will
produce blank sky spectral file which is to be used in 'xspec' to subtract
background, as well as background sky images.  Please see:


for details.

Also, we have released new GIS night earth event files.  All the night
earth data from 1993 June to 1999 August are accumulated to achieve
the exposure time ~13 Msec.  We expect these files would be useful
to estimate/subtract GIS non-X-ray background.  Please look at:


for more details.

*  Notice on the recent GIS gain correction from the GIS team          *

The following is the original announcement from M. Tashiro, A. Kubota 
(Univ.of Tokyo), H. Kubo (Tokyo Inst. Tech.), K. Ebisawa(NASA/GSFC ASCA-GOF) 
and the GIS team.  This is also posted at:


Please go to this page to see the figures cited in the text.

1. The radial gain correction formalism has been changed from 2nd order
    to 4th order polynomial function (GIS-S2).

The raw pulse-height channel (PHA) for each GIS event is converted
into energy channel (PI) through the following processes (1)-(3)
(see also http://heasarc.gsfc.nasa.gov/docs/asca/gain.html).

(1) Correct temperature dependence of the PMT gain
(2) Correct position dependence of the PMT gain
(3) Compensation of secular changes of the gain map

The gain map correction factors are provided through tables contained
in the "temp2gain.fits" file. The correction factor of (3) had been 
described with a 2nd order polynomial function of radius until now.
(The 3rd and 4th order factor has been put in place in the software but
has not been used.)

In figure 1, we show the center energy (PI) of the copper-K line versus 
the radius for GIS-S2 since the launch. Circles indicates the PI channels
calculated using the old 2nd order polynomial correction functions. 
The solid line in the panels indicate the nominal energy of the line,
and the each panel corresponds to half year terms, like April to September
1993 (1), October 1993 to March 1994 (2), ..., October 1998 to March 
1999 (12).  We see in the last three terms a systematic discrepancy. 
It becomes 0.8 % at most in the outer region ( 60 ch < r < 70 ch) of the
field of view for this half year, although the discrepancy is less than
0.5 % at 1CCD nominal position. Although we have to note that we have
systematic uncertainty of the GIS gain of +/- 1 % (region between dashed
lines), the residual structure indicates that the 2nd order polynomial
formulation no longer describes the radial trends well. Therefore we adopt
4th order polynomial descriptions and succeeded to compensate them in the
accuracy of < 0.5 % (figure 1 asterisks).

Please note again that the discrepancy of the Cu line energies for the
recent data expected by using the old (2nd order) gain map is at most ~0.8 %
(figure 1), which is smaller than the +/- 1 % systematic error.  Therefore,
Guest Observers will not have to reprocess their data for most practical
purposes. In the newly released table, we provide the 4th order coefficients
for the data since 1997 October. The GSFC/ASCA GOF will reprocess the data
after that occasion to revise the archival data as well as the newly
processed data but will not be re-distribute to observers. The old and new
correction factors are shown in figure 2.

	We have started using this new formalism in the processing of
	new data at GSFC.  Moreover, even though the impact to GOs is
	believed to be very small, we will reprocess all sequences obtained
	after 1997 September, in reverse chronological order.  We estimate
	that the reprocessing will be completed sometime in January, 2000.
	These reprocessed sequences will have version number 7.3.6 or later.

2. Azimuthal gain fluctuation at the rim region of GIS-S3

The GIS team warns that gradual degradation at the rim region (r > 23 
arcmin or  r > 90 channel) of GIS-S3 is causing an azimuthal gain 
fluctuation. The discrepancy is observed in the region outside 90 
channel from the detector center and could be up to 0.5 % nominally 
though we see double peak feature from the azimuthally integrated 
copper-K emission line obtained after 1998 July 1 (figure 3).

However, please note that as a default events from the detector region 
r > 88 channel are not assigned the sky coordinates  by "ascalin" and thus
truncated in the standard pipe-processing and "ascascreen".
Therefore this warning would apply only to the Guest Observers who
have reprocessed their data by running 'ascalin' by themselves 
in order to revive the event from the outermost region.