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

Rosat Status Report #86

                ROSAT Status Report # 86
                    March 1, 1994

ROSAT Satellite Status
ROSAT is presently in a test phase during which its behavior in a
stabilized operating mode is being studied. In this mode the satellite is
angular momentum stabilized if no starsensor information is available in
the shadow. In this case the magnetometers (the sources of the
irregularities) are not used by the control procedure. 

This stabilized mode has now been tested for almost a week and has proven
very stable. MPE decided to continue with the observing program starting
around midnight (UT) of Feb. 26/27, at that point the timeline switched
over to 5 days of PSPC observations. Re-starting the observing program with
the PSPC, MPE will be able to continue the monitoring of ROSAT's behavior
for another 5 days before they resume observations with the HRI on March 4.

Update on ROSAT naming conventions
This is a follow-up to ROSAT Status Report #76 which described the
conventions employed in naming ROSAT observations and processing
software versions.  That report contained the following diagram
(here updated) to illustrate the situation:

time     SASS 3_0         Rev 0         U.S. FITS           RODFITS    
 ||          |              |               |                (MPE)     
 ||          |              |               |                  |       
 ||      SASS 5_9           |               |                  |       
 ||    ------------ - - --------- - - ----------- 27 Jan 93 ---------
 ||      SASS 6_0         Rev 1             |                  |       
 ||          |              |               |                  |       
 ||          |              |               |                  |       
 ||      SASS 6_9           |               |                  |       
 ||    ------------ - - --------- - - ----------- ~1 Mar 94 ---------
 ||      SASS 7_0         Rev 2            RDF >---->+<-----< RDF
 ||          |              |                        |
 ||          |              |                       RDF
 ||          |              |                        |
 ||          .              .                        .
 ||          .              .                        .
 \/          .              .                        .

Standard Analysis Software System (SASS) versions used to process ROSAT
data are characterized as Revision 0 ("Rev0"), Revision 1 ("Rev1"), or
Revision 2 ("Rev2"). SASS versions 3_0 to 5_9, which were used prior to
January 27, 1993, are called Rev0; while SASS versions 6_0 through 6_9,
which were used from Feb. 1993 to Jan. 1994, are called Rev1; and SASS
version 7_0 and up, used after Feb. 1994, are called Rev2.

The new RDF data-product format (Rationalized Data Format, introduced 1994
March) unifies the different German and U.S. FITS formats into a single set
of formats.  All data processed prior to 1994 January (regardless of
whether processed with Rev0 or Rev1 SASS) will be REprocessed using the
latest Rev2 SASS software and formatted into RDF data products.
Observations performed after 1993 Dec 22 will be not require REprocessing
since they will be initially processed with Rev2 SASS software and
formatted directly into RDF data products. Hence all data in the public
archive will eventually be replaced with Rev2-processed, RDF-formatted data
products, resulting in a uniform public archive of all ROSAT observations. 

PROSCON update:
With IRAF/PROS version 2.3 (released 1993 December), output from the
task "qpspec" (in the xspectral package) has been changed to conform to
recommended RDF FITS standards. The source and background off-axis
histograms are now in files (*_soh.tab and *_boh.tab, respectively)
separate from the spectral data (*_obs.tab). To permit use with old
versions of XANADU/XSPEC, the new PROS spectral files can be converted
back to old-style spectral file format using the xspectral task
"downspecrdf".  This change initially led to some problems with
"proscon" (which required an IRAF/PROS work-around - if you need this
contact: "hotseat@cfa.harvard.edu"), but proscon has now been modified
to handle files produced by downspecrdf (so now you don't need the 
PROS patch).

A tar file containing the new version of proscon has been placed on
"legacy.gsfc.nasa.gov" anonymous ftp, under:


An alternative work-around for this file conversion is to use the IRAF task
stsdas.fitsio.stwfits to dump the PROS "qpspec" output files to FITS files
and then use ftools.caltools.stw2pha to make a FITS pi file for XSPEC. This
has the advantages of producing an OGIP FITS-format pi/pha file and
allowing conversion of HRI spectral files. This does not make a response
matrix. ROSAT FITS matrices are available on legacy.gsfc.nasa.gov under
/caldb/data/rosat/pspc/cpf/ and /caldb/data/rosat/hri/cpf/ (where cpf means
calibration product file). The disadvantage with the ftools route is that
it can currently only be used for sources within a few arcminutes of the
detector center (the folding-in of off-axis corrections is under
development and will soon be available). 

As noted in ROSAT Status Report # 74, FTOOLS are a collection of utility
programs to create/modify or examine data files in the FITS format. FTOOLS
form the basis of the ASCA analysis system, but there are a large number of
general FITS tools which are useful for the analysis of ROSAT data, along
with an increasing number of ROSAT-specific tasks. The FTOOLS tasks run
either standalone on all common platforms, and/or can be installed as an
IRAF package. At the current time it is recommended that the caltools
subpackage of ftools be run outside of IRAF to avoid memory allocation

The FTOOLS release version can be found as compressed tar files in the
/software/ftools/release/ directory of the anonymous FTP account on


In PROS 2.3, qpspec produces seq_obs.tab, seq_soh.tab, seq_boh.tab

1) If the source is within a few arcminutes of the detector center use:

STSDAS.FITSIO. STWFITS -> seq_obs.stwfits, seq_soh.stwfits,seq_boh.stwfits
STW2PHA-> seq_obs.pha (FITS)

(This route allows conversion of PSPC and HRI files. STW2PHA can be run
outside of IRAF or from FTOOLS.CALTOOLS, the former is highly recommended,
to avoid memory allocation problems - see above). 

2) If the source is off-axis use:

seq_obs.tab, seq_soh.tab, seq_boh.tab -> xray.xspectral.downspecrdf 
-> seq_prdf_obs.tab
proscon seq_prdf_obs.tab  
->seq_prdf_obs.pha  (SF)

(PSPC only, the output file is binned to 34 channels and is of old SF
format, and you must use the new version of proscon, see above) 

Comparisons between X-ray spectral fitting packages XSPEC and PROS
There are differences between PROS and XSPEC which have been traced to
differing assumptions in the default models.  In particular, results
obtained with XSPEC's "wabs" and "raymond" models are NOT directly
comparable with PROS.

However, meaningful comparisons between PROS and XSPEC, using the
Raymond model with finite absorption, can be made by tailoring XSPEC
to the abundance and absorption values used by PROS.  This can be done
by setting the XSPEC abundance ratio = 1.26 in versions  of XSPEC prior to 

Galactic absorption in both XSPEC and PROS is approximated as a piecewise
cubic polynomial.   Wabs in XSPEC and abs in PROS use identical
coefficients for the polynomials.  PROS determines the absorption based on
the central energy in a given PI bin.  XSPEC performes a different numerical
approximation when there is an edge within a given PI energy bin.  The
differences between the two models is only noticable in soft, highly
absorped spectra around the carbon edge.  The difference is only about 3%
which is less than the systematic errors in the PSPC response matrix at the
Carbon edge.