SPEX Version 2.0

There are several reasons for an update of the spectral code and analysis package. Some reasons are given here:

• Better atomic data have become available
• More transitions need to be included
• More flexibility for including more complex models (e.g. for photo-ionized plasmas)
• More flexibility and easier user interface
• More computer memory for high-resolution X-ray data
• Options for fitting grating spectra of spatially extended sources
• Options for fitting multiple regions of extended sources simultaneously, taking psf effects into account
• Improved description of the response matrix for X-ray grating data
• better portability to other platforms

Version 2.0 is now available

Currently (november 2004) the new version is operational at SRON and a few selected test sites. A few parts of the software need further documentation. The software has been installed succesfully on Solaris and Linux platforms.  There is also a Dec/alpha version, but that will not be updated frequently as we do not have direct access to such a platform. The update of the atomic parameters is still continuing. It has to be decided when that will be fully included in the new release.

Main differences with version 1.10

The main differences with version 1.10 can be summarized as follows:
 

1. Version 1.10 used fortran77, version 2.0 uses standard fortran90. In particular features like dynamical memory and derived data types are used frequently, as well as many other features of f90 (array and matrix manipulations, modules, etc.) The advantage of fortran90 is here the larger flexibility as compared to fortran77, while keeping the backward compatibility; as compared to C the better arithmathic crunching capacity of fortran is desirable (and as a survivor of the punchcard era, I am more familiar with fortran anyway).
2. the menu structure of version 1.10 has been translated into a command structure. This is in particular advantageous for complicated things like coupling of parameters among different spectral components etc.
3. it is now possible to fit spectra simultaneously when they are coupled spatially by finite psf effects
4. some new spectral models have been added (in particular for warm absorbers)
5. more spectral line broadening options (arbitrary profiles, and for spatial extent for grating spectra)
6. in the MEKAL calculations, wavelengths have been slightly improved according to recent work by Phillips et al. 1999 (A&AS 138, 381), and a few small corrections for some obvious mistakes have been made

Click the above link for the current release of this version.