[Isocops] [Ibexhigts] another background subtraction file
Nathan Schwadron
nschwadron at me.com
Tue Dec 7 20:40:34 EST 2010
OK, thanks Paul
This will be the basis of a new cull.
Nathan
On Dec 7, 2010, at 7:36 PM, Paul Janzen wrote:
> Not really sure who all's on this list, but here goes...
>
> Two files are necessary to handle good/loose culls for IBEX-Hi: the file
> listing good times (typically HiCullGoodTimes) and the file listing
> backgrounds (frequently HiCullBGCRAN). This email is about the second
> file: numerous variants are in the SVN repository, and I've just gone and
> added another, as promised in September.
>
> There are two backgrounds which we subtract at present: the cosmic ray
> background, and "loose" cull offsets. The cosmic ray background can be
> treated two ways: either as a constant 57 mHz, which is what we've been
> doing for the most part, or as a background that changes per orbit, either
> scaled by neutron monitors or calculated by the ratio of qualified triples
> to unqualified triples. The latter should be more accurate. The former is
> less work and more conservative.
>
> If the CR bg is calculated by the ratio of triples on IBEX-Hi, then both
> the loose cull offset and the CR background can be calculated directly from
> IBEX data and the good times file, and do not need to appear as magic
> numbers (with the exception of orbits 55 and 56, and a few early orbits
> where the spin rate was crazy). For the time being, though, magic numbers
> in files fit into the processing scheme, so that's what I'm using. The
> ISOC will eventually generate these numbers itself.
>
> The list of background files in idsup/culldata so far is:
> HiCullBGCR : includes only cosmic ray bg at a constant 57 mHz; for use
> with good times but not useful with loose culls.
> HiCullBGCRAN : includes 57 mHz cosmic rays and Additional Noise -- i.e.,
> the loose culls. This is the one I have been updating as new loose
> culls come in.
>
> HiCullBGCRvar : an old attempt to incorporate a variable CR background
> based on unqualified triples. It goes to orbit 76 and does not
> supply uncertainties. It also does not handle loose culls.
>
> HiCullBGCRAN-2009-11-30
> HiCullBGCRAN-2010-02-15
> HiCullBGCRAN55
> HiCullBGCRAN.pj :
> old versions of HiCullBGCRAN that I don't think anyone is using.
>
> The one I've added is:
>
> HiCullBGCRANvar : includes a variable CR background (based on unqualified
> triples) and the loose cull offsets, as calculated by a program from
> the IBEX-Hi data. Also includes sigma, much of which is systematic as
> opposed to statistical. I will keep this one up to date too.
>
> As per what I presented at the September meeting, I believe (barring
> errors) that HiCullBGCRANvar is what should be used for the most accurate
> heliospheric data. It's basically HiCullBGCRAN but handling the CR
> background "better"; it also does a slightly better job on early loose
> culls because it no longer assumes (as the original pass through orbits
> 31-34 did) that the additional isotropic noise is of constant brightness in
> time. If you use HiCullBGCRANvar, though, bear in mind that it will change
> all your numbers back through orbit 14 or so, somewhat. And in the
> interests of full disclosure, the second background test is not
> reassuringly consistent with this variable background, (although a few other
> things are).
>
> If you want to see what happens, set
> hi_bg_file
> in your scripts to
> /usr/local/ibex/ops/config/idsup/culldata/HiCullBGCRANvar
> and let'er rip. Probably hi_bg_file is set to HiCullBGCRAN now.
>
> If not, this additional file need not change your life.
>
>
> For reference, I note that there are a bunch of files
> ops/config/idsup/HiCull*
> which are obsolete but kicking around, too. Their format is somewhat
> different from the present bg file...
>
>
> Paul Janzen.
>
> Please restrict discussions on this email list to non-ITAR sensitive topics.
> _______________________________________________
> Ibexhigts mailing list
> Ibexhigts at lists.sr.unh.edu
> http://lists.sr.unh.edu/mailman/listinfo/ibexhigts
More information about the Isocops
mailing list