README FOR SIS BACKGROUND FILES

This directory contains SIS background event files: 

s0bgd_all.evt	s1bgd_all.evt
s0bgd_06.evt	s1bgd_06.evt
s0bgd_08.evt	s1bgd_08.evt
s0bgd_10.evt	s1bgd_10.evt
s0bgd_12.evt	s1bgd_12.evt
s0bgd_14.evt	s1bgd_14.evt

These files contain events from 3 deep field observations without 
any obvious sources and with a total exposure of about 120 ks. The 
events were cleaned using the "select mkf" command in XSELECT 
(with ELV_MIN > 20 and SAA = 0). Hot and flickering pixels were 
removed using the "sisclean" command in XSELECT (sisclean method 
2, with cell size 5, log probability -5.24, background level 3). Finally, 
any spurious events which remained were removed by manual 
cleaning based on the light curve (i.e. excluding spikes and drop-
outs). 

The background rate is a weak function of the lower limit of cut-off 
rigidity (COR) used, except at the extreme ends (below 6 and above 
14 where the dependence is strong).

In order to prevent the extraction of astrophysical information, the 
files have a spurious pointing direction of RA =0, dec=0. This means 
that X=DETX and Y=DETY. TIME is set to DBLE(PHA) and the events 
are sorted in ascending order of DETX

To use these files in XSELECT, Guest Observers should choose the file 
which best matches the COR range in their source events file and 
read it in using the "read events" command. Next, extract a 
background spectrum. The best (and easiest!) way to do this is to use 
the same region filter file as was used for the source spectrum, 
making sure that the spatial binning and coordinate system (DETX, 
DETY) are the same. In XSPEC, read in the source spectrum with the 
"data" command, and the background spectrum with the "backgrnd" 
command. 


Accuracy/limitations:

SIS background consists of four components: extragalactic, Galactic 
and instrumental (of which hot/flickering pixels are treated 
separately).

(1) Hot/flickering pixels: The "sisclean" procedure removes the vast 
majority of flickering pixels. However, low duty cycle flickering pixel 
events cannot be cleaned out. Moreover, how many of the low level 
flickering pixels are left depends on observing conditions (Sun angle, 
CCD temperature etc.), the exposure time and the cleaning criteria 
used. The background event files should be a reasonable indication of 
the flickering pixel background, but not completely accurate.

(2) Extragalactic X-ray background: This component dominates the 
SIS background in the range 2-6 keV, where it is well-determined 
and relatively independent of pointing directions. At lower energies, 
the different Galactic Nh does make the determination unreliable. 
Note also that there could be unresolved, or barely resolved 
background sources in the background and/or target field of view. 

(3) Galactic X-ray background: No attempt has been made to create 
background event files that would enable subtraction of this 
component, as it depends entirely on the pointing direction. This 
component is very important around 1 keV. 

(4) Internal background (excluding hot/flickering pixels): The 
background event files should give some indication of the overall 
level of this component. However, its contribution depends on the 
observing conditions and cleaning criteria, as well as the clocking 
mode of CCD. The total level is around 0.008 cps/CCD in 4-CCD mode 
(which is the case for the background event files), 8% higher in 2-CCD 
mode and 15% higher in 1-CCD mode. This component is dominant at 
high energies, and has fluorescent lines of Al, Fe, Ni and Au. The line 
energies may be shifting over time.

The total background rate is about 0.05 cps per CCD chip (or 0.014 
cps per 3.5 arcmin radius circular beam), and the use of the 
background files will allow estimations accurate to ~0.005 cps per 
chip (0.0015 cps per 3.5 arcmin beam). At the lowest energies, this 
procedure leaves uncertainties due to the variable absorption of 
extragalactic background, the Galactic background, and the variable 
efficiency in removing flickering pixels. At the highest energies, the 
instrument (non-flickering) component leaves uncertainties 
particularly where it has lines.

For questions, please write to Koji Mukai: 

mukai@lheavx.gsfc.nasa.gov