Background Extraction with srctool

[JosephHall]

Hello,

I'm extracting spectra for nearby clusters (0.02 < z < 0.03) and I'm having issues with extracting background spectra. When I extract them at the same time as the source spectra (using BACKREG), the background spectra are often empty or have far fewer events than I would expect:

Additionally, the scaling values in the headers of these spectra seem off -- BACKSCAL and REGAREA are both 0 for the background. The srctool command used for this was:

Code: Select all

eventfiles=output/l500_testing_rebin/erass_demo0/evtool_erass_demo0_srctool.fits suffix=.fits todo=SPEC srccoord=fk5; 11:44:29.55, +19:48:53.18 clobber=yes lcemin=0.5 lcemax=2.0 extpars=5748.54549598445 srcreg=Regions/r500//Sourceerass_demo0.reg prefix=output/l500_testing_rebin/erass_demo0/srctool_tests/20240815_140532/srctool_erass_demo0_fullband_ backreg=Regions/r500//Backgrounderass_demo0.reg gtitype=FLAREGTI exttype=TOPHAT psftype=NONE

If, however, I extract the background regions separately (by setting them as SOURCEREG and BACKREG=None) then the spectrum is extracted normally with a much more reasonable number of events. This is for the same exact background region as above:

This was produced with:

Code: Select all

srctool eventfiles=output/l500_testing_rebin/erass_demo0/evtool_erass_demo0_srctool.fits suffix=.fits todo=SPEC srccoord=fk5; 11:44:29.55, +19:48:53.18 clobber=yes lcemin=0.5 lcemax=2.0 extpars=5748.54549598445 srcreg=Regions/r500//Backgrounderass_demo0.reg prefix=output/l500_testing_rebin/erass_demo0/srctool_tests/20240815_140342/srctool_erass_demo0_fullband_/BKG backreg=NONE xgrid=100 gtitype=FLAREGTI exttype=TOPHAT psftype=NONE 

My source and background regions for these plots are, respectively and including background point sources:

Code: Select all

circle(11:44:29.55,+19:48:53.18,2467.275726876526")
-circle(11:47:58.94140447123732,+20:13:11.159557856494189,60.0")
-circle(11:41:16.26309590446425,+21:56:25.036744948915555,60.0")
...

Code: Select all

-circle(11:44:29.55,+19:48:53.18,5748.54549598445")
circle(11:44:29.55,+19:48:53.18,6000.67279357512")
-circle(11:47:58.94140447123732,+20:13:11.159557856494189,60.0")
-circle(11:41:16.26309590446425,+21:56:25.036744948915555,60.0")
...

Additionally, some of the other clusters I'm considering at higher redshifts (z~0.08), the regions are smaller and the inner radius of my background annulus is closer (in angular dist) to the source region. For these cases, the background spectra are extracted fine, much more like the second image above. The headers for these backgrounds also have non-zero values for the BACKSCAL keyword, unlike what I recover for the nearby example.

Is there any reason for this difference in the extracted spectrum? And is there anything I've missed that can make srctool produce results closer to what I'd expect when doing combined source and background spectra? If not, what calculations are needed to evaluate the scaling keywords in the headers of these files?

Thanks in advance for all your help.
Joe.

[JeremySanders]

Hi Joe

Yes, it's because the background spectra are accumulated during the good time intervals (GTIs) of the source. The manual says

Please note that by default any background region is extracted using the same GTIs as the source. If the source region is large, this means that only a limited part of the background region data will appear in the background region outputs. In addition, an ARF appropriate for the source region is unlikely to be appropriate for the background region if they are large and/or separated. Therefore, the user is advised to extract large background regions separately and do a joint spectral fit on the source and background spectra. When modelling spectra, please note that the BACKSCAL includes effects such as loss due to the field of view, which are also included in the ARF effective area. The appropriate area if modelling using an ARF is likely to be REGAREA (the pure geometric area), whereas if modelling without an ARF (e.g. particle background), BACKSCAL is likely appropriate.

If the background region is large, then there will be periods when the background is visible when the source is not, and will not be included in the background spectra. Therefore if you want to do get all the data from the background, you need to extract the spectrum separately.

The reason for this is for point sources, when you want to have the background observed at exactly the same time as the source.

Please see above text from the documentation for the scaling of the background spectra keywords.

Jeremy