Come analyze HEASARC, IRSA, and MAST data in the cloud! The Fornax Initiative is now welcoming all interested beta users.
Next: polrot: Rotate polarization Up: crossarf: mixing model for Previous: crossarf: mixing model for Contents
Multiple emission components in a region
A region may contain more than one emission component whose spatial distributions differ – for instance an extended source together with a point source. Because the two components scatter into the neighbouring regions by different amounts, a single set of cross arfs is not enough; each component is instead handled as a separate source (see the model command), with its own arfs loaded by the arf command. Two cases arise:
- A component that is itself spread across several regions and mixes between them is given its own crossarf source. If there are two or more such components with different spatial distributions, each is a separate crossarf source and their cross arfs are distinguished by a source tag in the XFLT keyword name, “CrossArfSrcSFromNToM:arffile”, where S is the source number and, as before, arffile gives the contribution to region M from region N. The original “CrossArfFromNToM” form, with no source tag, is treated as belonging to source 1. The “CrossArfRegion” keyword is not source tagged, since the region a spectrum belongs to is the same for all sources.
- A component confined to a single region does not need a crossarf at all: it is added as an ordinary source with the arfs giving its contribution to each detector region.
Example 1: an extended source in both regions and a point source in one of them. The extended emission fills both regions and mixes between them, so it is a crossarf source (source 1). The point source lies in region 1 only; it is seen in both detector regions but does not mix, so it is an ordinary source (source 2) with the two arfs giving its contribution to each detector. Load the two spectra as two data groups and attach both sources' responses:
XSPEC> data 1:1 region1.pha 2:2 region2.pha XSPEC> response 1:1 region1.rmf XSPEC> arf 1:1 ext_1to1.arf XSPEC> response 1:2 region2.rmf XSPEC> arf 1:2 ext_2to2.arf XSPEC> response 2:1 region1.rmf XSPEC> arf 2:1 pt_1to1.arf XSPEC> response 2:2 region2.rmf XSPEC> arf 2:2 pt_1to2.arfThe extended crossarf (source 1) is the only mixing source, so its cross arfs use the plain (untagged) keyword form. In the spectrum for region 1:
XFLT0001 = 'CrossArfRegion:1' XFLT0002 = 'CrossArfFrom2To1:ext_2to1.arf'and in the spectrum for region 2:
XFLT0001 = 'CrossArfRegion:2' XFLT0002 = 'CrossArfFrom1To2:ext_1to2.arf'Define the two models:
XSPEC> model 1:ext crossarf*(tbabs*apec) XSPEC> model 2:pt tbabs*powerlawThe extended emission generally differs between the two regions, so give its data-group-2 parameters their own values (untying them from data group 1). The point source is a single object, so leave its data-group-2 parameters linked to data group 1 (the default for a new data group). The point source appears in region 2's spectrum only through its arf pt_1to2 – no mixing keyword is needed for it.
Example 2: two mixing components. If instead two components
are each spread across both regions with different spatial distributions –
for example a cool and a hot plasma phase – then each needs its own
crossarf, and the source tag distinguishes their cross arfs. Load
each source's diagonal arf as in Example 1 (arf 1:1 cool_1to1.arf,
arf 2:1 hot_1to1.arf, and so on), and in the spectrum for region 1:
XFLT0001 = 'CrossArfRegion:1' XFLT0002 = 'CrossArfSrc1From2To1:cool_2to1.arf' XFLT0003 = 'CrossArfSrc2From2To1:hot_2to1.arf'Then define one crossarf per component:
XSPEC> model 1:cool crossarf*(tbabs*apec) XSPEC> model 2:hot crossarf*(tbabs*apec)Here CrossArfSrc1... supplies the cool component's cross arfs and CrossArfSrc2... the hot component's; each crossarf mixes only its own component, and XSPEC sums the source contributions in every spectrum.
Next: polrot: Rotate polarization Up: crossarf: mixing model for Previous: crossarf: mixing model for Contents