Thursday, December 30, 2004


Majczyna et al. calculate X-ray spectra of iron-rich bursting neutron stars. This is an extension of an earlier paper dealing with H-He atmospheres.

Valluri et al. study the gravity wave signal from spinning-down pulsar.

Fargion considers some less obvious effects of the Boxing Day Tsunami including a possible LIGO signal.

Bekenstein sketches a new non-Einsteinian relativistic gravity theory which tends to MOND and Newton in the appropriate limits.

Keywords: neutron stars, gravity waves, MOND

Obscure fakeit problem

Craig spotted that when faking a background file with Gaussian errors I failed to correct the VARBKG array for the ratio of BACKSCALE keywords. This meant that the STAT_ERR column would be wrong if the source and background files had different values of BACKSCALE. Fortunately, this is unlikely to have occurred because Gaussian errors are only used with non-imaging detectors. Fixed as bug 11.3.1s.

Updated on Jan 7, 2005 to fix a typo in the original file.

Keywords: xspec

Monday, December 27, 2004


Calabretta shows that the HEALpix spherical projection is a member of a previously undocumented infinite class. He gives projection equations and FITS formalism.

Keywords: mathematical methods

TIMEREF keyword in lightcurves

Uddipan from TIFR reports that extractor sets the TIMEREF keyword to 'local' regardless of its value in the input event file. Removed the relevant calls from W_FBLC since this keyword is set to the correct (propagated) value by WSTDKY. Extractor v4.50.

Keywords: heasoft, extractor

Wednesday, December 22, 2004


Barwick et al. summarize the prospects for neutrino astronomy and cosmology.

Keywords: neutrino

extractor update for region files with ', d, or r sizes

Ziqin noted a problem with using region files output from fv which have region sizes given in arcminutes (ie with an ' appended to the number). Checking the ds9 manual page on regions
I see that appending d for degrees and r for radians is also allowed. I've updated extractor to support these (v4.49).

Note that I'm not yet supporting appended p (for physical) or i (for image). I also don't support all the possible region shapes. I should fix that one day.

Keywords: heasoft, extractor

xselect.mdb minor change

Added entries for ROSAT:catnum and ROSAT:lststr. These were being specified at the instrument level (eg ROSAT:PSPC:catnum) but this led to irrelevant warning messages when reading in the data files. One day I need to overhaul this so that it is not necessary to define all the parameters at the mission level.

Keywords: heasoft, xselect

Tuesday, December 21, 2004

rtsmpr left out of bug fix tar files

Oops, in the fixes for 11.3.1o and following I included the new xscf.h and tclout.c that call rtsmpr. I corrected for this by adding rtsmpr.f to the tar files for the o-r bugs. People who picked up the bugs tar file between Dec 6 and 21 may get a missing symbol problem.

UPDATE: and of course I need to include the new FIT/Makefile as well to ensure that rtsmpr.f is compiled.

Keywords: xspec

g95 v4.0.0

Installed the GNU f95 compiler alpha release from binaries available on the Mac OS X High Performance Computing site. These installed in /usr/local and included some of the latest gcc libraries and include files. These should not be picked up by default because /usr/local/bin is near the end of my path. The fortran compiler is called gfortran. The HPC site notes that it may require the -lstdc++ flag to recognize some standard functions. Also some libraries may require sudo ranlib -s. I may have to update gcc to make C and Fortran programs play together.

Keywords: software, fortran


Marshall et al. discuss using Bayesian evidence to compare datasets and ask whether they were generated from the same model system.

Keywords: Bayes, statistics

Monday, December 20, 2004

Sporadic problem of large error bars in faked spectra

David Buote has had occasional problems when the first fakeit in a session produces a spectrum with some large error bars and hence small chi-squared. I was unable to reproduce the problem locally however inspection of a file that David produced shows that the STAT_ERR column is large for channels that are in groups but are not the start of the group. These channels should have zero counts and errors. Only the former is true. Inspection of the wrtfak.f code shows that the error array written is not initialized to zero unlike the count array. I've passed instructions back to David on how to fix this bug and now await the results of his tests.

UPDATE: David reports that the bug is fixed so placed the correct version of wrtfak on the
xspec bugs page as 11.3.1r.

Keywords: xspec

Better diagnostics when reading data into xspec

Added the filename to diagnostic messages generated when reading data into xspec. I ran into the problem that I couldn't tell whether an error message was due to a problem with the source data file or its background file.

Keywords: xspec

Friday, December 17, 2004


Markevitch et al. find a supersonic shock in Abell 520. This is the second clear example of such a shock.

Franceschini et al. present the results of Spitzer, Chandra and optical follow-up of the ELIAS-N1 field. Of 122 X-ray sources with fluxes > 1e15 cgs IRAC detects 102. 39% of Chandra sources are type I AGN, 23% type II, the rest show starburst or normal galaxy characteristics although all are AGN.

Kemball & Martinsek investigate bootstrap resampling to assess the fidelity of radio interferometry imaging.

Keywords: clusters, AGN, statistics

dmextract defaults to Gehrels errors

Thanks to a question from Bing Jiang (in China ?) I've noticed that in CIAO 3.1 (July 2004) the default behaviour of dmextract was changed so that STAT_ERR is set to 1+sqrt(N+0.75), Neil Gehrels' approximation for low count data. I disagree with this strongly since it overestimates the errors for reasonably bright sources. The old behaviour can be recovered by psetting error=gauss. The POISSERR keyword is also set incorrectly - being set to F for sqrt(N) errors and true otherwise.

UPDATE: Note that this doesn't matter for xspec normally because if POISSERR=T then sqrt(N) errors are assumed however running the spectrum through grppha will change the keyword to F so xspec uses the errors in the STAT_ERR column.

Keywords: CIAO


Mateos et al. analyze X-ray spectra of 1137 serendipitous XMM sources. They find absorption in only ~40% of the number of sources predicted by XRB synthesis models.

Lee & Ravel investigate the feasibility of measuring ISM grain composition due to EXAFS in the Fe K absorption edge observed using XRS on Astro-E2.

Konacki et al. confirm a 3-day period transiting giant planet in OGLE-TR-10.

Keywords: papers, AGN, EXAFS, planets

Thursday, December 16, 2004

compbb problems

Made some progress on a long-standing bug report from David Ballantyne. He found an instance where an error command failed when a new minimum was found then a negative flux model was generated. Managed to reproduce this in a simple script with particular values of the compbb parameters. Passed information onto Ken Ebisawa to see whether he has any ideas.

Also, noticed that if the energy range limits are changed but the blackbody and electron temperatures are not then the model is not updated properly. This bug probably never arises in practice. Reduced the number of calls to CMPBBK by a factor of 2. Checked in these changes but will wait for feedback from Ken before putting anything on the bugs page.

UPDATE: added to bugs page as 11.3.1p. Ken confirmed that David's problem was because the approximation used is only valid for kTe >> kT and his fit had kT > kTe.

Keywords: xspec, bug


Mike Garcia and colleagues use Chandra and HST to find a possible 3 x 10^35 erg/s source coincident with the central black hole in M31.

Fujita et al. use hydro simulations to predict high resolution X-ray spectra observed with Astro-E2.

Asai et al. do MHD simulations of a galaxy moving through the ICM and show that a magnetized medium is required to produce the cometary tail seen in Chandra observations of A2125.

Nicastro et al. estimate the mass density of the WHIM using results from X-ray absorption spectroscopy and show that it is consistent with the predicted value.

Keywords: papers, black holes, clusters, WHIM

Wednesday, December 15, 2004

PLT problem with GRID X

Frank Marshall came up with an instance where "GRID X # #" had no effect. After exchanging e-mail with Allyn we established that :
1. For multi-window plots "WINDOW ALL" followed by "GRID X" does not work.
2. "GRID X" doesn't work if the X-axis is logarithmic.

Remember to contact Allyn after he returns to work on Jan 4 to get a new version of PLT.

Keywords: lheasoft, plt

Interpretations of Probability

I should make sure I understand this article from the Stanford Encyclopedia of Philosophy since I pontificate on the subject often enough.

Keywords: probability, Bayes


Sanderson et al. look at XMM and Chandra observations of Abell 478. Chandra continuum kT is higher than that from XMM - this effect is minimized if isothermal regions are chosen indicating difference is due to spatial variations in temperature and the different spectral responses of the instruments. They find four hot spots which they interpret as shock-heating due to the radio source in the cluster center.

Cassano & Brunetti calculate thermal and non-thermal emission from clusters using an evolutionary model.

Georgantopoulos & Georgakakis find no evidence of X-ray absorption in an XMM sample XBONGs.

Mauche et al. discuss Fe-L shell diagnostics and identify density diagnostics.

Keywords: papers, clusters, AGN, X-ray spectroscopy

Removing case changing from openwr

Bryan and I decided to remove the code from OPENWR that automatically changes names of new files to lowercase. This was originally intended to ease transfers of files between VMS and Unix systems and as such is obsolete. It has produced problems in xselect when the user has a mixed case directory tree. Files changed are src/xanlib/sysdep/sys.*.

Keywords: lheasoft

Up and Running

This is a test. This is only a test. If this were a real post it would be marginally more interesting.