XRT News (Image and Extended Source Analysis)
Last modified July 31, 1996
This document lists various information regarding the XRT instrument/calibration, and image and extended source analysis with ASCA.
Guest Observers who have questions regarding the XRT instrument and image analysis are encouraged to notify the ASCA GOF by sending e-mail to ascahelp@athena.gsfc.nasa.gov.
Current Contents
- Release of new XRT responses (July 31, 1996)
- Analyzing extended sources with ASCA data (February 15, 1994)
Release of new XRT responses (July, 1996)
ascaarf v2.62, as well as the XRT response files, xrt_ea_v2_0.fits and xrt_psf_v2_0.fits, was released on July 1996 through ASCA GOF Previous versions of ascaarf (v2.53 and earlier) combined with older XRT responses (xrt_ea_v1_1.fits/xrt_psf_v1_1.fits and earlier), has been known to have the following problems:
- The GIS Crab spectrum was not perfectly fitted with a power-law, and 5 % residuals remained.
- SIS and GIS did not give consistent normalizations. SIS normalization was considered to be more realistic, and the GIS gave 15 - 20 % smaller normalizations.
- The GIS Crab spectrum gave the photon index 2.15, which is a little higher than generally believed values (2.08-2.12; e.g., Toors and Seward; Ast.J.79,995).
- The GIS Crab flux was 18 % smaller than the standard flux
(e.g., Toors and Seward; Ast.J.79,995).
These problems were corrected in ascaarf v2.62 and xrt_ea_v2_0.fits/xrt_psf_v2_0.fits. Essentially, two changes have been made:
- XRT effective area (xrt_ea_v2_0.fits) and the point spread function (xrt_psf_v2_0.fits) are revised based on a new Nagoya ray-tracing code. The gold optical constant ( ) was tuned above 2.7 keV so that the residuals of the GIS Crab spectral fit become minimum. The PSF was re-calibrated based on the new GIS Cyg X-1 observations, resulting in a slightly wider PSF and 13 % decrease of the effective area within a 6 arcmin radius from the image center (hence increased GIS flux).
- Only modifying the XRT calibration files did not solve the problems shown above
satisfactorily,
hence additional correction filters are introduced so that the GIS Crab spectrum can be
fitted with a power-law, and the four detectors give the same normalization.
Please note that ascaarf v2.62 has to be used with xrt_ea_v2_0.fits and xrt_psf_v2_0.fits to make the latest and most reliable ARFs. The correction filters are included in the ascaarf v.2.62 source code, and can be turned off by specifying the hidden parameter "arffil = no". By turning off the ARF filter and combining with older XRT calibration files (xrt_ea_v1_1.fits and xrt_psf_v1_1.fits), ascaarf v2.62 is able to make the same old ARFs as were made with ascaarf v2.53 and earlier.
Please click here for more detailed comparison of the old and new responses.
As you may know, the point spread function (PSF) of ASCA's X-ray
telescopes (XRT) has a relatively sharp core (FWHM of 50 arcsec) but
broad wings (half-power diameter of 3 arcmin). In addition, the GIS has
its own PSF which is comparable in width to that of the XRT but of a
different shape (a Gaussian with a sigma of 30 arcsec at ~6 keV and a
1/sqrt(E) dependence).
For analyzing point sources, the practical consequence of the broad PSF
is that a generous extraction radius should be used to capture most of
the counts (e.g., 4 arcmin for the SIS, 6 arcmin for the GIS).
For extended sources, however, the astrophysical implications of the
broad PSF are more complicated and harder to deal with. If you are
analyzing an extended source, then please note the following effects
which, if not taken into account, could lead to spurious scientific
results. It is important to realize that these effects are not
calibration uncertainties or instrumental defects, but features
intrinsic to the design of the XRT which emphasizes effective area as
much as angular resolution.
The XRT PSF has a sharp core and broad wings. If your extended source
has a brightness distribution which is peaked on the same scale as
the PSF, then the outer parts of the image will contain a significant
proportion of counts from the core. For example, the moderate
redshift cluster A2218 has a core radius of about 1 arcmin. If the
image is divided into annuli (3 arcmin wide centered on the core)
then the second annulus (between 3 and 6 arcmin) will contain more
emission from the cluster core than from the parts of the cluster
which actually lie within the annulus.
The effect described above will be compounded by telescope vignetting
if the core of your object is close the optical axis of the XRT.
The XRT scatters high-energy X-rays more than low-energy X-rays. In
other words, the PSF broadens with increasing energy. One consequence
of this effect is to introduce spurious (outwardly increasing)
temperature gradients in isothermal distributions. And if there is a
negative temperature gradient present, its size will appear reduced.
If an off-axis region is specified and an ARF constructed for it then
if this region really contains counts from closer to the optical axis
then the ARF will not be correct and the difference will be in the
sense of increasing the measured temperature.
Unlike the SIS, the GIS has an intrinsic PSF which is of comparable
size to that of the XRT. The energy dependence of the GIS PSF is a
further complication.
In view of these effects, extracting spatially resolved spectral
information from ASCA observations is not trivial, especially in the
case of sources which are of order 10 arcminutes or less in size. It is
likely, therefore, that extensions to present analysis techniques are
needed to address the issue.
The ASCA instrument teams are calibrating the PSF as accurately and
completely as practicable. To this end, the ASCA Team recently conducted
a series of supplementary calibration observations of the bright X-ray
binary Cygnus X-1. Images extracted from these observations, which
placed the target at various positions in the field of view, are
available for several energy bands in the directories:
They are accessible via anonymous FTP at heasarc.gsfc.nasa.gov. The
ray-tracing PSFs which have been available since March 1993 have been
moved to:
In general, dealing with the effects of the broad PSF will involve
deconvolution, the complexity of which will depend on the particular
case in question. Modifications are being made to XSPEC that will
provide the potential for dealing with this problem by allowing
models to read information from the headers of PHA files. The GOF
staff will be available to advise users in generating the appropriate
deconvolution software for other, more complex cases.
The ASCA Team is also producing a PSF generating FTOOL called MAKE_PSF
which will output a PSF for arbitrary energy and position. MAKE_PSF will
be included in the next-but-one FTOOLS release in Spring or early
Summer.
Charles Day and the ASCA GOF
Analyzing extended sources with ASCA data (February 15, 1994)
If you have any questions concerning ASCA, visit
our Feedback form.