 |
 |
Hello folks, following the observation from Daniele http://www.slac.stanford.edu/cgi-bin/lwgate/VUB-RECOIL/archives/vub-recoil.200209/Author/article-70.html that the mm2 problem could be tested looking at the events with low mm2 in MC, I started scanning individual events in the cocktail that had mm2<1.5 GeV. I realized that I had a complete misconception of those events: if one missed a particle, the measured mm2 is on the positive side: if originally emiss=pmiss (missing neutrino), with a missing particle emis+epart>=pmiss+ppart If one adds a particle mm2 becomes negative: if originally emiss=pmiss (missing neutrino), with a particle in eccess emis-epart>=pmiss-ppart so far, we got the right expectation ... Problems arise when one mistakes the mass of a particle. In the case of a charged track, if you underestimate the mass of a particle (i.e. call pion a kaon), the momentum is correct, but the energy is underestimated and therefore mm2<0. In the case of a neutral cluster, if one mistakes a neutral hadron for a photon, the energy is correct, but the momentum is overestimated and therefore also in this case mm2<0. In conclusion, contrarily to what was intuitive to me, the low mm2 tail is due to underestimating the mass of a particle. There are therefore two sources of low mm2: fake particle reconstruction and low mass assignment. As far as the first one is concerned I just found out that we are keeping also tracks which have no Dch hits, but that are reconstructed as several GeV tracks. This can be seen in http://www.slac.stanford.edu/~rfaccini/phys/vub/aug02/noDch.eps where the scatter plot of the theta and the pt of the tracks with no Dch is shown. It looks like throwing away tracks which have pt>300 MeV and no SVT info would be healthy. As far as incorrect mass assignment is concerned it looks like the biggest effect is due to neutrals and we are back to neutral hadron identification ... The difficult part to understand is why data and MC disagree and in particular why the effect is more visible on MC than on data. talk to you soon ric