This section contains a partial recipe for a single observation that has a bright extended source in the FOV. It is assumed that you have already applied the recipe from the previous section to extract the raw image of the full FOV, and to extract the spectrum of the entire FOV. Because that spectrum contains the emission from the bright source, the contribution due to the residual soft proton contamination or SWCX may be entirely overwhelmed by the source spectrum. Thus, we need to extract spectra for a region that minimizes the contribution from the bright source.
For the full FOV, one should have followed the recipe in the previous section up to “BREAK POINT 2”. In the analysis directory, you should have a number of subdirectories: ffov_mos1S001, ffov_mos2S002, ffov_pnS003_0, and ffov_pnS003_4.
It is assumed that you have constructed region files, similar to the reg*.txt files of the previous recipe that exclude the bright source, perhaps named reg*-annu.txt. Using these region files, re-execute from “BREAK POINT 1” to `'BREAK POINT 2'`, placing the output products into the subdirectories annu_mos1S001, annu_mos2S002, annu_pnS003_0, annu_pnS003_4, and annu_spectra. Then fit the spectra for all the Galactic components, the residual soft proton flare emission, and the SWCX emission. We can now use the fit parameters to construct the residual soft proton flare images and the SWCX images using the recipe in the previous section with the following exception:
At “BREAK POINT 3” we need to convert the normalization of the residual soft proton emission from the region that was extracted for the spectral fit (annu) to the region that was extracted for the full FOV (ffov). This is done with sppartial, where “xxx” is the normalization from the fit to the soft proton emission. (It does not make a difference whether it was a power law or a broken power law.)
sppartial fullimage=ffov_${M1}/${M1}-fovimspdet.fits fullspec=ffov_${M1}/${M1}-fovt.pi regimage=annu_${M1}/${M1}-fovimspdet.fits regspec=annu_${M1}/${M1}-fovt.pi regnorm=xxx_mos1 sppartial fullimage=ffov_${M2}/${M2}-fovimspdet.fits fullspec=ffov_${M2}/${M2}-fovt.pi regimage=annu_${M2}/${M2}-fovimspdet.fits regspec=annu_${M2}/${M2}-fovt.pi regnorm=xxx_mos2 sppartial fullimage=ffov_${PN}_0/${PN}-fovimspdet.fits fullspec=ffov_${PN}_0/${PN}-fovt.pi regimage=annu_${PN}_0/${PN}-fovimspdet.fits regspec=annu_${PN}_0/${PN}-fovt.pi regnorm=xxx_pn
This routine will return the normalization that is required for the spectra in “ffov”. We can now return to the “ffov” directories and run the recipe after “BREAK POINT 3”. Note that no changes are required for creating the SWCX images.