Monday, February 28, 2005

error in lstat fitting

Fixed error in lstat fitting spotted by Craig - dtconv was used in ab_lstat before being defined. Probably a mistake introduced when changing areascal and backscal to vectors.

Keywords: heasoft, xspec

Saturday, February 26, 2005

extractor and NULLs

Fixed extractor so that it rejects events whose TIME column contains a NULL. Note that only the TIME column is tested for NULLs.

Keywords: heasoft, extractor

Wednesday, February 23, 2005


Bryan & Voit use numerical simulations of cluster formation to argue that much of the ICM cools at high redshifts hence any heating mechanism to provide the extra entropy must operate at high redshifts as well.

Ballantyne & Fabian analyze XMM-Newton observations of the BLRG 4C+74.26 and find evidence for a relativistic Fe line extending into the ISCO for a maximally rotating BH.

Sesana et al. compute the expected GW signal from coalescing MBHs and predict ~90 events above S/N=5 in a three-year LISA mission.

Keywords: Clusters, AGN, Gravitational Waves

Tuesday, February 22, 2005

xselect Chandra response generation

CIAO 3.2 introduces a new tool, mkacisrmf, for building ACIS RMFs. I modified the xselect Chandra response generation script (xsl_chandra_acis_makeresp) to look for mkacisrmf and use that in preference to mkrmf.

Note that this script doesn't test whether the input file used the appropriate gain file for mkacisrmf.

Keywords: heasoft, xselect

Monday, February 21, 2005

Lightcurve output from extractor/xselect

Martin Still requested a new option for xselect/extractor. Currently the lightcurve output writes times relative to the start bin (ie starting at zero) with the offset from spacecraft time given by the keyword TIMEZERO. Martin wanted the option to have TIMEZERO set to zero so that it is easier to combine lightcurves from different observations. This is now implemented using the extractor boolean parameter lctzero - set this to no for TIMEZERO=0. Within xselect give the argument offset=no to the extract lightcurve command.

Keywords: heasoft, xselect, extractor

Thursday, February 17, 2005


Realized that dummyrsp with channel offset and size specified requires that the number of energy bins equal the number of channels. If this is not the case then a seg fault will occur. So, improved the help to make this clear and modified dmyrsp.f to force this condition if the user tries to violate it.

Keywords: heasoft, xspec

Wednesday, February 16, 2005


Motl et al. calculate the integrated SZE effect for a sample of theoretical clusters with different assumed baryonic physics. They show that the integrated SZE provides a very accurate estimate of the mass.

Yao and Wang detect NeIX, OVII, and OVIII absorption lines in Chandra spectra of XRBs and use these to study the large-scale hot interstellar medium of the Galaxy.

Quirrenbach summarizes current technological progress on space-based coronographic imaging to find planets.

Keywords: clusters, ISM, planets

Wednesday, February 09, 2005


Phil Uttley and colleagues show that the rms-flux relation in AGN and XRB implies a non-linear, exponential form for their light curves. The variability process must be multiplicative (not shot-noise).

Higdon et al. use Spitzer observations of a field surveyed by the VLA and Chandra and conclude that optically-invisible radio or X-ray sources are not generally detected at 24 microns implying that they are unobscured AGN.

Jones and DeYoung perform MHD modelling of relic radio bubbles in the ICM and confirm that magnetic fields can stabilise the bubbles giving them lifetimes of 10^8 years.

Tuesday, February 08, 2005

Contour plots and xhistory

A couple of small xspec bug fixes.
11.3.1u The label with minimum and contour levels was being overwritten in plot contour.
11.3.1v The xhistory fit package was not being written out. I think I commented this out of lvmrun.f when we were moving code to v12. Only the JPL folks use this at the moment.
Updated this on Feb 9 by moving wfthis call from lvmrun to lvmfit so it occurs after the fit parameter object is set.

Keywords: xspec

Thursday, February 03, 2005

Seminar: John Baker on MBH mergers

John gave a lunchtime PSU simulation and sources talk remotely from Goddard. He summarized the interest in observing massive black hole mergers using LISA. Vecchio has recently argued that at least six months of observations are required to constrain important binary MBH parameters. This constraint requires that planned LISA sensitivity be improved at the low frequency (<>1e6 Msun) black holes at high (z>5) redshifts. If we can also get physical constraints from the merger/ringdown then this requirement is eased.

John also summarized recent numerical GR work. The Lazarus collaboration is working on fourth-order differencing - an approach that should provide much improved accuracy for a given amount of computation. Initial work is promising. The Goddard group is working on adaptive mesh methods to calculate the GW signal in the far-field. The most recent calculation is of a head-on BH-BH collision where the outer boundary of the calculation is at 256 M with steps in the resolution at 8, 16, ... M.

Keywords: Black holes, GR, LISA


Pizzolato and Soker consider the analogies between cavities in PNe and in X-ray images of clusters and suggest that it may be possible to find binary black holes based on the symmetry of the cavities created.

Duane Liedahl looks at the effects of resonant Auger destruction on Fe K lines from compact sources.

Sturrock et al. give statistics based on combining two or more power spectra.

Keywords: Black holes, X-ray spectroscopy, statistics

Constellation-X matrices

I created the latest iteration of the Constellation-X response matrices. All effective area numbers were supplied by Paul Reid (CfA) in Excel files (ipg_020105.xls and opg_020105.xls). He included all detector filter and QE efficiencies. There are two basic options. The In-Plane Grating (IPG) is the old baseline while the Off-Plane Grating (OPG) is the new proposed, higher-resolution option. The two options cover different regions of the mirror so the calorimeter effective areas also differ.

The calorimeter (XMS) response matrices assume a 2 eV FWHM resolution which is independent of energy. Effective areas were supplied from 0.25 - 10 keV and the responses were created with 0.0003 keV energy bins.

The IPG RGS matrices are for three orders with resolutions (FWHM) of 0.087, 0.0435, and 0.029 Angstroms, respectively. All three matrices cover 5 to 50 Angstroms with energy binning of 0.01, 0.005, and 0.003 Angstrom, respectively.

The OPG RGS matrices are also for three orders with wavelength-dependent resolutions supplied by Flanagan and Rasmussen (in files FR_OPG_res_m#.dat). The first order matrix covers 5 - 50 Angstroms with binning 0.003, the second order 5 - 38 with 0.0015 binning, and the third order 5 - 25.5 with binning 0.001.

All matrices have COMMENT keywords in the primary header which describe the input (and were lifted from Paul Reid's Excel files).

Keywords: Constellation-X

Wednesday, February 02, 2005


Fixed a bug in genrsp for dispersive instruments whose resolution varies with wavelength and is entered via a file. As far as I know this option has never been used before.

Keywords: heasoft