PK's Crab Spectra


Date: Mon, 17 May 2004 17:59:41 +0200
From: Peter Kretschmar 
Subject: OSA 4 validation - Crab spectra stability
To: Carol Anne Oxborrow , Juhani Huovelin , Niels Lund , Niels Jørgen Westergaard , Søren Brandt , Stefan Larsson , Sami Maisala , Carl Budtz-Jørgensen , Jerome Chenevez , Silvia Martínez Núñez , Paul Connell 
MIME-version: 1.0
Content-transfer-encoding: 7BIT


  Dear all,

  while my script compilation checks are ongoing, I've found
a bit of time to look at the Crab pointings from rev 170 which 
I've previously analyzed with my integration account installation which should
correspond to OSA4 (/home/isdc/pkretsch/Cons/obs/Crab0170/).

  Doing some XSPEC fitting with the following constraints:
- limit to range 5.0-40.0 keV
- fixed powerlaw slope (2.1) and NH (0.2E22, i.e. insignificant)
I've noted that most (but not all) of the pointings forming the 
outer rim of my selected dithering pattern had a 'funny' spectrum 
with a bump around 15 keV or channel 160, while the other spectra
seemed more or less similar. You can see this in a rough
way by looking at the following plots:

* All spectra of the pointings at the rim:
  Crab_jmx1_rim_en.pdf

* All spectra of the 'inner' pointings:
  Crab_jmx1_inner_en.pdf

Naturally one will have to look closer at the details, but
in general the spectral flux determination seems quite 
stable and reasonable up to 15-20 keV in the 'inner' zone.
The outer zone, in contrast, not only has much more scatter
in the reconstructed fluxes but also this bump.

  Going one step further, I have built a summed spectrum
from all 'inner' pointings and tried to fit that. The first
result is that around 5-10 keV the slope seems to be close
to the canonical 2.1, so I've fitted the data there
(NH=0.2, slope=2.1, norm=13.7 [too high]) and then plotted
the whole spectrum, which now deviates below 5 and above 10 keV
based on the response instance 0011:
  Crab_jmx1_inner_sum_PL2.1.pdf
Allowing for a free PL slope gives the non-canonical values
of NH=0.2, slope=2.26, norm=19.0 and the following plot:
  Crab_jmx1_inner_sum_PLfree.pdf

  Note: I did expect to find a 'wrong' spectrum, since these
responses were tuned to the interim j_src_spectra. But it
is somewhat interesting to see where the deviations lie.

  Just for fun, here's the average spectrum of the 'outer'
or 'rim' pointings fitted again between 5 and 10 keV with
a fixed slope first:
  Crab_jmx1_rim_sum_PL2.1.pdf

  I hope this is useful for a start and am ready to discuss
further tests and choices. For JEM-X 2 I may take a look 
tomorrow, but will have to see if I find the time.

  Regards,
  Peter