SIS Background - Local Background and Blank Fields
- New: a list of SIS blank fields observed during 1996-2000 (see below).
- Cosmic X-ray background - sky position dependent.
- Non X-ray (particle) background - time dependent.
- Hot and flickering pixels - time dependent.
This figure shows the spectrum of SIS background. In upper panel, composite spectra (for 8 SIS chips) of blank sky observations are shown after removal of the hot and flickering pixels. Also shown are the models of the particl background for the 8 chips, as derived from observations of the night Earth. Lower panel shows the net Cosmic X-ray background. Taken from the SIS Background article in ASCAnews no 2.
Because the SIS field of view is relatively small, the detector position dependences (e.g., the vignetting of cosmic X-ray background) are generally not important. (One exception to this is certain areas of SIS-0 chips 2 & 3, where a light leak causes increased flickering pixel rate particularly when the line of sight is close to the Bright Earth.)
The particle background dominates above ~6 keV; between about 1.5 and 6 keV, the dominant source of background usually is the cosmic (extragalactic) X-ray background, while at lower energies, Galactic background becomes more important. And at lowest energies, flickering pixels can make a contribution. See the SIS Background article in ASCAnews no 2 for more details.
For a circular extraction region of 3.5-4.0 arcmin radius (which might be typical for point sources), the average background is roughly 0.02 counts per second per SIS. The average background integrated over a single CCD chip on one SIS is roughly 0.05 counts per second. With either the local or the blank sky method (see below) background can typically be estimated to about 10-20% accuracy. You should compare these numbers with the brightness of your source to determine how careful you need to be.
Note, however, that the relative contribution of the background to the total (source+bgd) is a function of energy. For example, even for a 10 counts per second X-ray binary, background subtraction may well be necessary to determine the photon index accurately, since the particle background makes a measurable contribution near 10 keV where the SIS has a low sensitivity to X-rays.
When a careful background subtraction is desirable, we recommend trying multiple methods (and/or multiple blank sky files) to estimate the sensitivity of your results to the background subtraction method.
1. Local background subtraction
In this method, you simply find a source-free region of the SIS in the observation of interest. The advantage is that all the time dependences are automatically factored in; in particular, this method is suitable for obtaining a background-subtracted light curve of relatively faint sources. If only the time-average background is subtracted, the net light curve will retain the time dependence of the particle background. Also, this method correctly estimates the Galactic X-ray background, although it does not account for the point-to-point fluctuation in the cosmic background on the arcminute scale.
The disadvantage is that there is little room in the SIS for a truly source-free region. This is true of 1-CCD mode observation of point sources; for a 1 counts/s source the entire CCD chip may clearly be dominated by the source counts (i.e., the shape of the point spread function can be traced to all 4 corners of the chip). For a fainter source, say at 0.2 counts/s, an area roughly equal to the source extraction region may qualify as source-free.
2. Blank Sky Background method
In this method, background is extracted from the same region of the detector as the source region, but from different observations. We have constructed composite blank sky event files from several PV phase observations; these are available from Calibration Database
Unfortunately, the SIS background depends on the CCD clocking mode, the event threshold and/or lower energy discriminator setting, and the epoch of observation (as the radiation damage degrades the sensitivity of the SIS). Thus, reliance on the PV-phase blank sky data above (which are taken in 4-CCD mode, with the original event threshold of <0.4 keV) is not desirable.
There are only a limited number of blank field observations, however, which do not cover the parameter space (clocking mode, epoch of observaation etc.) well. It is highly unlikely that we can devise an automatic process which works reliably, unlike the GIS blank sky background. Instead, we provide a table of SIS blank fields, currently limited to the latter half of the mission (1996-2000), so users can make their own assessment of the limitation of the PV phase blank sky data. We plan to expand this table to cover the entire mission in due course.
This file was last modified on Monday, 30-Jul-2001 22:37:44 EDT
HEASARC Guest Observer Facility
If you have any questions concerning ASCA, visit our Feedback form.