Contributors: Alex Dzierba, Richard Jones
September 18, 2004
The starting point for this study is the results from Radphi for the resolution of the lead-glass detector in the environment of a bremsstrahlung photon beam, as reported in technical note Radphi 2003-701. The principal results are as follows.
| [1] | |
| [2] |
Using these resolution factors, Alex has generated some Monte Carlo events for GlueX photoproduction and smeared them for energy and position errors. His results are summarized here. In comparing measured and Monte Carlo mass resolutions, it should be noted that the results depend on the reaction in which the particle appears. For example, a π0 that appears in the decay b1->ωπ0 would have a different energy distribution in the lab than one from direct π0 photoproduction. Alex has also explored the dependence of his result on the t-slope parameter and observed a dependence of about 1 MeV in the width of the π0. Within the range of these uncertainties, the agreement between Alex's Monte Carlo resolutions and both the Radphi and E852 observations is good, suggesting that the Radphi resolution results describe both detectors.
As a more complete check on the lead-glass mass resolution for GlueX, I have done a complete simulation and shower reconstruction analysis of the following channels, assuming a bremsstrahlung spectrum in the range 8-9 GeV.
The above mass spectra are the result of all reconstructions that result in the correct (generated) number of photons, with no check on whether they are nearby the original gammas in position or angle. They are essentially what the real detector and analysis would produce for these reactions under realistic conditions. The mass resolutions verify the results from Alex based on the simple Monte Carlo smearing technique. The plots also show that significant amounts of combinatoric background can be present in reconstructed mass spectra.