From GlueXWiki
Jump to: navigation, search

Present: M.D., A.D., J.S.

  • General:
  • M.D:
    • Studied running sum uncertainty and concluded that the present determination (as done in the proposal version provided for the Gluex review) is correct.
  • A.D:
    • Studied running sum uncertainty. After including an overlooked bin-size term in his simulation, found consistent results and now agrees with Mark's handling of the uncertainty.
  • S.Š: Simon could not be at the meeting but provided a written report:
    • Instead of the NMC parameterization of g1p(x,Q2) which Alexandre gave me, and another one of F1p(x,Q2), and instead of using Callan-Gross to obtain F2p and Wandzura-Wilczek to obtain g2p etc. I have implemented a newer and quite stand-alone parameterization of g1p and g2p which I need in my polarized Bethe-Heitler code. The formulas, if someone is interested, are in that paper.
    • I have found three mistakes in their paper and there were some iterations with Simula after he sent me his fortran code through which the errors could be isolated and fixed. Now I have nice and smooth g1p and g2p which, according to the authors, should be valid for x >= 0.02. In the Bethe-Heitler code smaller x than that are occasionally required, but I think that the present parameterization is much more suitable for our purposes than the old one. See these two figures for g1p and g2p for the heat maps of these two functions.
    • Another gain with the new g1p and g2p is also that the resonant part, not just the DIS region, is now nicely covered: these are all the groovy structures seen in the plots. However, this makes the calculation much slower, even though it is just a parameterization: for each x and Q2, there is a loop that goes through all 4* resonances and all sorts of parameters and there is no way to speed this up by, say, "fitting a fit". This also has the potential to make the calculation in the main BH code unstable:a typical integrand for integration over the final-mass-squared now inherits all these wiggles: see this picture. To solve this I tried (and I am still not quite satisfied with it) to find a compromise between fast+imprecise+stable integration of these wiggles and a slow+precise+unstable one ... This is a quite painful exercise but I have something now that is sort of operational and does not crash.
    • Here are the results. This plot shows BH observables for Egamma = 12 GeV, assuming p_e- = p_e+, with phi_e- = 0, phi_e+ = 180, as a function of theta_e- = theta_e+. This was my usual convention to display results as there is so much kinematic freedom in specifying d^6(sigma)_BH. The left panel shows the elastic BH cross-section (dashed curves) and the sum of elastic+inelastic (full curves). The right panel shows the asymmetries in the same notation. For p_e- = p_e+ = 5 GeV/c the asymmetry becomes positive and large and looks weird at first sight... But have a look at this plot which shows the full polar angle range. Similar features are present at all momenta, they are just being pushed towards smaller angles. And the abrupt termination of the inelastic component apparently must have a clear reason: the pion production threshold. So it all makes some sense.
    • If Mark's simulation indicate that BH rate is comparable to the hadronic rate --- pending unresolved trigger issues hinted at by Justin --- I have more oil to add to the fire: what about BH from nuclei in the target (which was also on my menu)? So far I've been dealing only with protons.
    • Regarding the endorsement letter for which I asked Marc VdH and Vladimir Pascalutsa, I have received a response from Vladimir, and he is less than enthusiastic: "As for the endorsement, i am willing to help.. but cannot take the initiative, too many things on my plate already ..and as I said before, the physics case of sum rule violation is not very appealing to me. i can join a group and edit some bits and pieces. Maybe you could try to involve younger ppl, e.g., Franziska Hagelstein. Hopefully Marc has a good idea on who could lead this effort."
    • But Marc has not responded yet. Unless you have other suggestions, I might wish to ask Franziska for help, but it won't be of the same caliber as if the two "big guys" wrote the letter.