'''Observing GRB in X-rays:''' 

<ul>

<li>GRB and Background Basics</li>


'''Spectral Slopes'''

'''GRBs'''  You should look at [[The_Great_Bruce_GRB_Information_Library#General_GRB_Observation_Topics|GRB General Observing>Slopes]]  explanation of GRB spectral slopes; I use here BAT-like instrument results.  Sakamoto+11 says BAT sees 
 low-E PL w/Photon index -1.6 for LGRB.    For SGRB, sakamoto says low-E Photon index -1.2 for SGRB, very different results from Nava 

'''The DXRB'''  diffuse X-ray Background is complicated and modern papers point out all sorts of curvatures, etc.   Me, I'm a lazy bum and use very very old parameterization Zombeck p. 197 , http://ads.harvard.edu/cgi-bin/bbrowse?book=hsaa&page=197, weirdo units are KeV cm^-2/s/Sr/keV 

NA              ~ 0.1 keV <E<5 keV        (To remind you again: Zombeck parameterization does not cover the soft bump.)

8.5*E^(-.4)   5 keV <E<25 keV

167.*E^(-1.38)   5 keV <E< ~ MeV

(Better parameterizations have a low-E exponential cutoff and all kinds of curvatures, e.g. I(E) = 7.9 E-0.29 exp[-E/41.1 keV] keV cm-2 s-1 sr-1 keV-1 from ~ 5-30 keV. http://www.slac.stanford.edu/econf/C0307282/lec_notes/kahn/kahn1.pdf)

Here are my plots and discussions of how this plays out for choosing energy bands for observations.  (It is from my script singu!~/projects/ubat/xethresh.pro). 

[[File:grb_vs_dxrb_spectra.png]]

So, the DXRB is almost as hard as SGRB up to ~ 25 keV, then it is very steep. 

Now let's see that in Counts space, which helps visualize results below. 

[[File:grbvdxrb_counts.png]]

                  <li> X-rays-Observing>Energy Band Selection</li>
The DXRB has a "thermal-ish" bump that dominates from tenths of keV to ~ 3-5 keV.   For this reason, you don't want to go below 5 keV, so '''5 keV is the obvious lower bound of energy'''.  What do you gain by going to high energy? 

If you integrate spectra in count rate space, you see that LGRB rise briskly  to ~ 25 keV, then gradually flatten.  
[[File:intgd_spectra.png]]

However, if you make a VERY SIMPLE assumption that the background is dominated by the DXRB (it isn't!!  There are particles!!!!)  you could make a simple argument that you should integrate to high energies because S/N keeps improving: 
[[File:simplified_snr_vs_Emax.png]]

So, I suspect the particle background spectrum is very hard, and truth is not so optimistic (without active veto).  Still, it seems that it should be OK to go all the way up to 100 keV (but recall Chris' caution about being saturated by particle and other events just over your maximum energy threshold; take care with finite resolution and dynamic range). 

Note that the SGRB are still rising at the highest energy while the LGRB is flattening. 



                  </ul>