Up to top level
AO15   AO16   AO17   AO18   AO19   Backgrounds   Calibration   Conference   Data   Docs   EPICMOS   EPICpn   Feedback   Gallery   Misc   OM   Pending   PhD_Theses   Publications   RGS   RadMonitor   SAS_Hardware   SAS_WS   SASv16.0   SASv16.0_Installation   SASv16.1   SASv16.1_Installation   SASv17.0   SASv17.0_Installation   SASv18.0   SASv18.0_Installation   SciSim   Simulators_other   Suggestions   Trash   Visibility   XMM-bouncing   XMM-news   XRPS   XSA   esas   incoming  

Logged in as guest

Viewing EPICpn/8707
Full headers

From: bianchi@fis.uniroma3.it
Subject: Extraction regions with holes
Compose reply
Download message
Move To:
2 replies: 1 2
1 followups: 1

Private message: yes  no

Notes:

Notification:


Date: Fri, 6 Jun 2003 08:30:03 GMT
From: bianchi@fis.uniroma3.it
To: xmmhelp@xmm.vilspa.esa.es
CC: bianchi@fis.uniroma3.it
Subject: Extraction regions with holes
Full_Name: Stefano Bianchi
Submission from: (NULL) (193.204.162.2)


In order to correct the pileup in some sources, I have tried to extract spectra
from a region with a hole in the center. I have tried this trick on two sources
(one extremely affected by pileup, the other marginally) and adopted for the pn
regions with outer radius of 60 arcsec and holes with radius 5,10,15,20 arcsec.
Then I produced rmfs and arfs with rmfgen and arfgen for each spectrum.
The resulting fit parameters seem to go in the right direction, producing
steeper photon indexes, more compatible to past X-ray observations than the
spectrum from the complete region.
However, the fluxes are probably screwed up.
The source flux is inversely proportional to the hole radius: it can be as high
as four times that of the complete region when you use a 5 arcsec hole, twice
with 10 arcsec and significantly higher even with 20 arcsec.
It is clear to me that pileup affects count rates, making them lower, but I
don't think the effect can be so high. Moreover, it is not clear to me at all
that it should also affect the flux in the same simple way, lowering it.
I think the higher flux I'm measuring is indeed an artefact of a wrong arf or
something like that.
How much small can the extraction region be in order to be flux calibrated? This
should also apply to a hole, shouldn't it?

Thank you for your help,

Stefano


Reply 1

Resend
From: Nora Loiseau <xmmhelp@xmm.vilspa.esa.es>
To: bianchi@fis.uniroma3.it
Subject: Re: Extraction regions with holes (PR#8707)
Date: Mon Jun  9 08:57:33 2003
Dear Stefano,

here I append the reply about your problem by EPIC expert. He
suggests that you send the data for him to analyse them in detail.
You can send them to my private account (nloiseau@xmm.vilspa.esa.es)
with FITS files attached.

Best regards,

Nora
-------------------------------------------------------------
  the PSF normalisation in arfgen for circles and annuli uses
the EXTENDED (King) model and for 1 keV photons from an on-axis
source should give the correction factors:

Annulus   encircled energy factor
0 - 60      0.92
5 - 60      0.70
10 - 60     0.44
20 - 60     0.20

for the EPIC-pn camera (the MOS gives similar values).

So, if there were no counts at all in the inner 5 arcseconds, due to
pile-up, then the flux of the 5-60 arcsec annulus should be 0.92/0.7 =~
30% higher than that of the 0-60 arcsec circle. A factor of four is
not possible.

If the source area is more complicated than an annulus (.e.g. it
is an annulus with circles removed to account for other sources)
then another, less reliable, PSF correction has to be used.

Please send the 0-60 and 5-60 spectra and I'll take a detailed look.



Reply 2

Resend
From: Nora Loiseau <xmmhelp@xmm.vilspa.esa.es>
To: bianchi@fis.uniroma3.it
Subject: Re: Extraction regions with holes (PR#8707)
Date: Fri Jun 13 14:39:20 2003
Hi Stefano,

here is the reply by Richard Saxton:

------------------------------------
   thanks for the files, they have provided the answer. There is an
easy fix for the user in this case.

The annuli used to extract the data are not quite right. They are
composed of an included and excluded circle but with slightly
different centres, e.g. the 10-40" annulus is made up from:

CIRCLE  2.589350E+04  2.588070E+04  8.000000E+02
!CIRCLE  2.588550E+04  2.587830E+04  2.000000E+02

Unfortunately arfgen considers this to be two separate circles rather
than an annullus and applies a PSF correction which is known to have
major problems.

The simple solution is to ensure that these two circles have the
exact same centre, either with fv or by re-extracting the spectra.

I'll make sure that this is well documented in the next release.

Regards,

Richard





Followup 1

Compose reply
Download message
Date: Fri, 13 Jun 2003 16:50:19 +0200
From: Stefano Bianchi <bianchi@fis.uniroma3.it>
To: Nora Loiseau <xmmhelp@xmm.vilspa.esa.es>
Subject: Re: Extraction regions with holes (PR#8707)
Thanks Nora (and Richard!).
I'll try this fix.
However, if the two centres are different, it is probably an xmmselect 
bug, because I imported the region from ds9, simply changing the radius 
of the circle, without changing the centre. So, there's probably a sort 
of approximation of the coordinates when importign the region from ds9 
to xmmselect.
I will try this, too.

Thanks,

Stefano

Nora Loiseau wrote:
> Hi Stefano,
> 
> here is the reply by Richard Saxton:
> 
> ------------------------------------
>    thanks for the files, they have provided the answer. There is an
> easy fix for the user in this case.
> 
> The annuli used to extract the data are not quite right. They are
> composed of an included and excluded circle but with slightly
> different centres, e.g. the 10-40" annulus is made up from:
> 
> CIRCLE  2.589350E+04  2.588070E+04  8.000000E+02
> !CIRCLE  2.588550E+04  2.587830E+04  2.000000E+02
> 
> Unfortunately arfgen considers this to be two separate circles rather
> than an annullus and applies a PSF correction which is known to have
> major problems.
> 
> The simple solution is to ensure that these two circles have the
> exact same centre, either with fv or by re-extracting the spectra.
> 
> I'll make sure that this is well documented in the next release.
> 
> Regards,
> 
> Richard
> 
> 
> 
> 
> 
> 



Up to top level
AO15   AO16   AO17   AO18   AO19   Backgrounds   Calibration   Conference   Data   Docs   EPICMOS   EPICpn   Feedback   Gallery   Misc   OM   Pending   PhD_Theses   Publications   RGS   RadMonitor   SAS_Hardware   SAS_WS   SASv16.0   SASv16.0_Installation   SASv16.1   SASv16.1_Installation   SASv17.0   SASv17.0_Installation   SASv18.0   SASv18.0_Installation   SciSim   Simulators_other   Suggestions   Trash   Visibility   XMM-bouncing   XMM-news   XRPS   XSA   esas   incoming  

Logged in as guest


Please make your (short) question the subject of your request!


Web interface using JitterBug ... back to the XMM home page