Subject: Re: JEMX spectral bumps

    Hello Niels,

    Worried a little bit of declining JEM-X 2 spectrum from Rev. 0044
I extracted JEM-X spectra for almost all Crab observations. The results
are shown in the attached figure, where spectra are plotted against Crab
canonical model. There are: 0043 (black), 0044 (cyan), 0102 (blue), 0170
(magenta), 0239 (red) and 0300 (green). Spectral extraction was done with
all parameters set to default values, for scws with off-set < 3 degrees.

    There is a clear discrepancy observed for Rev. 0102 spectrum, which
has much lower absolute normalization. "5 keV line" feature is seen in all
spectra except for 0300, but the position of this bump seems to move
towards lower energies for later observations, i.e., 0170 and 0239. Almost
all spectra, with 0102 and 0170 exceptions, show slope somewhat steeper
than the canonical value 2.1. Al last, the most recent data (0239,0300)
exhibit strong excess below 4 keV.

    If you want to see the logs or other files, they are in:
/unsaved_data/scratch3/lubinski/crab/obs/0102_3deg_jmx2_OSA5
and similar subdirectories of /unsaved_data/scratch3/lubinski/crab/obs.

    The new IC files you mentioned will be for OSA 5.1?

     Best wishes,

            Piotr



Niels Joergen Westergaard wrote:
> Hello Piotr,
> 
> thanks for your analyses.
> 
> The bump that you have disclosed in your may be related to the ARF.
> Since the response of the instrument at lower energies is not
> well known because of the mix of various effects such as gain changes,
> event rejection criteria and the low high voltage, then the Crab
> calibration has been a 'real' calibration i.e. the ARF has been measured
> by the response to the Crab spectrum.
> 
> I have probably been too keen on ironing out in the 'features' in the
> spectra used for calibration (the On-axis Crab observations) so that
> statistical fluctuations may play an unduly large role in the ARF
> determination. The ARF for the REV 170 JEM-X1 plot that you give does
> indeed show a peak around 5 keV.
> 
> Then the action on me is to make sure that narrow features are not
> reflected in the ARF and to include more Crab observation i.e. also
> the ones that have off-axis angles up to, say, 3 degrees to follow
> your example.
> 
> The result of this will a new version of the RMF IC files.
> 
> When that is ready I hope that you will have time to evaluate the
> situation again.
> 
> Thanks for pointing this out.
> 
> Regards
> 
> Niels Jørgen
> ----------
> Niels J. Westergaard
> Danish National Space Center
> Juliane Maries Vej 30, DK 2100 Copenhagen O
> Phone: +45 35325705; FAX:  +45 35362475
> 
> 


-- 
Piotr Lubinski
INTEGRAL Science Data Center    Phone:  +41 22 3792 146
16, Chemin d'Ecogia             Fax:    +41 22 3792 133
CH-1290 Versoix                 www:    http://isdc.unige.ch/