Monday, April 30, 2007

hk file extension numbers in xselect

Change requested by Suzaku XIS team to allow multiple extensions in HK files. The read hk command now allows an extension to be given eg filename+3. This entire name plus extension is passed to maketime when doing a select hk. The code change does in principle allow the same syntax to be used when reading an event file however this is disabled because in some places xselect appends a +1 extension descriptor to the event file name.

Thursday, April 12, 2007

xspec v11 and gfortran block data problem

Put in workarounds for the block data initialization problem with gfortran. None of these should have any discernible effect but we shall see. The changes were in the default values for genfit, the default names for model components, the default gain parameters in response, and the atdata.f routine used by the NEI models.

This latter has been changed in xspec v12 as well.

Errors on fluxes of model components

A frequently requested enhancement for xspec is to be able to estimate the error on the flux from part of the model eg the model without Galactic absorption. I've now written a more detailed explanation and provided a helpful Tcl script. It should be possible to incorporate this functionality within xspec itself.

Monday, April 09, 2007

gfortran compiler and BLOCK DATA

The gfortran compiler appears to have a known issue with intializing common variables in a BLOCK DATA. This leads to problems in xparse with the XPRSBD block and causes msetup to fail. The workaround is to create a dummy routine which calls XPRSBD and insert a call in any xparse routine that uses variables in the BLOCK DATA. This fixes the problem and xspec v11 builds and runs the session/xspec.xcm script successfully. However, there are other BLOCK DATA in HEAsoft so we need to make similar fixes for them. Note that in Fortran 90 and later code BLOCK DATA is supposed to be replaced by MODULE.

Wednesday, April 04, 2007

xspec background models

I wrote a slightly longer explanation of the v12 replacement for v11 /b models. I also added a link from the xspec FAQ.

Preprints

Evans et al. describe the on-line database of Swift GRB lightcurves.

Tombesi et al. find a variable redshifted Fe K line with a modulation timescale of 27 ksec using XMM observations of NGC3783.

Bonavita & Desidera note from a literature search that there appears to be no difference between planet detection rates in single and binary star systems.

Migliori et al. analyze XMM observations of the radio lobes of Pictor A and finds that the particle energy is above the equipartition value and that spacial changes in radio flux are due to magnetic field variation.

Zhang et al. discuss how to test standard gravity on cosmological scales by comparing matter overdensity to gravitational lensing.

Kurucz argues that the exoplanets detected so far are "cryptoplanets" formed from matter ejected by protostars.

Bryan et al. (the CMU astrostatisticians) present a new method to determine multidimensional confidence intervals and apply it to the cosmological data.

Porter & Ferland discuss the addition of He I like lines to Cloudy and their use for diagnostics.

Afshordi considers the accuracy of mass estimates from SZ surveys of clusters.

Erickcek et al. note that the X-ray quantum calorimeter (XQC) rocket flight can be used to place strong constraints on interaction cross-sections between baryons and dark matter.

Rudnick et al. point out that the WMAP cold-spot corresponds to a dip in the number counts of NVSS sources. If this is Sachs-Wolfe effect it would require a void of radius 140 Mpc.

Werner et al. analyze XMM and Suzaku observations of Sersic 159-03 and find a central soft excess but no evidence for lines ruling out thermal sources.