STIX spectrogram FITS files are not natively compatible with NASA’s X-ray Spectral Fitting Package. However, work is ongoing such that this powerful, free software can be used with STIX spectra, should sswidl’s OSPEX prove insufficient (ex., OSPEX can only fit with chi-squared minimization and one spectrum at a time).
This example demonstrates the results of fitting a broken power law with coronal abundances to a STIX spectrum.
Input
Input files are named identically to those accepted by OSPEX: a FITS file containing the spectrogram, with multiple rows and the file name extension .fits, and a temperature response matrix with the file name extension .srm, that is referred to in the spectrogram FITS header.
The spectrogram and response function are shown below.
Zooming in and looking at the index shows that for the flare peak from 17:24:30 - 17:25:30, the correct row has index 13 (indices in XSPEC are from 1 not 0).
Data is then loaded into XSPEC with the row specifier in curly braces, for example:
xspec.AllData("1:1 stx_spectrum_20210908_1712.fits{13}")
The spectrum of the single time bin is shown below.
Specifying Coronal Abundances
XSPEC does not natively support solar coronal abundances, only photoshperic. Like in the NuSTAR example, specify the abundance file manually.
xspec.Xset.abund=f"file {abundance_file}"
Model Set-up
The chosen model is a broken power law. By default the break energy is 5 keV, which is low for solar flares, so initially it can be set higher. Here it is set with an initial value of 15 keV, increments in 0.1 keV and a maximum of 18 keV. This parameter will automatically be frozen.
m1=xspec.Model("bknpower")
m1.setPars({2:"15.0 -.1,,,18"})
Fitting once over an energy range of 8-30 keV results in a chi-squared statistic of ~ 2000. This can be improved by a second fit, now letting the break energy parameter vary freely, by unfreezing it before fitting again:
p2=getattr(m1.bknpower,'BreakE')
p2.frozen=False
The chi-squared statistic of the fit is now a much more reasonable 40. Note that initially fitting with the default break energy parameters will result in a very low break energy and a higher chi-squared, instead of reaching the better solution provided by fitting twice.