2007 October, HXD Team
update: 2007 November 27th
PIN and GSO background with longer exposure was uploaded.
Note: The original background files have relatively short exposure of ~90 ks,
while the new ones have 800 ks for the PIN and 490 ks for the GSO and thus significantly higher statistics.
The Poisson errors of the original files are smaller than the systematic
errors of the background. So it is not necessary to redo your simulations using the new
background in most cases, though the significance of signals may increase a little bit.
However, we recommend to redo simulations, ONLY if you are
following the simulation recipe in section 5.5.2 of the technical
description AND if you use the absolute values of the chi^2 as
criteria of your observation.
The HXD covers the highest energy range of the Suzaku band-pass, from 10 to 600 keV. The detector is a collimated well-type phoswich array designed to achieve the lowest background, and therefore the highest sensitivity ever achieved, in this band.
In spite of the low detector background, the sensitivity of the HXD is dominated by the systematic error of the instrumental background estimates. We suggest you fully take into account this error in all simulations.
To restate: only objects brighter than the systematic error of the background estimate at the energy band in question are detectable.
The background systematic error will have an energy dependence as well as a time dependence. See Technical Description for details. For example, with 100 ks exposure, the systematic errors are +/-3% both for PIN in the 15-40 keV band and for GSO in the 50-100 keV band.
Please note that we do not have separate rmf and arf files for the HXD.
The PIN response is roughly 10% higher when a point source is observed at the "HXD nominal" pointing position, compared with the default ("XIS nominal") position. The GSO response files have identical effective areas at these pointing positions, although the off-diagonal elements are slightly different.
Please also note that the background in the above files drop sharply above ~77 keV for the PIN, and ~750 keV for the GSO. These are outside the nominal bandpass of these detectors, and should not affect the feasibility studies.
For an accurate feasibility study, systematic uncertainties in the bacground estimation must be taken into account. This can be done by reading the background files into xspec both as background and correction files, and then using the cornorm command. See Chapter 4 of Technical Description for details.