Optimization of Jet Energy Resolution and Response of the D0 Detector

Abstract

The D0 Experiment (E-740) is located at the Tevatron of the Fermi National Accelerator Laboratory (Fermilab). As one of only two general-purpose experiments situated to take full advantage of $p\overline{p}$ collisions at a center-of-mass energy of 1.8 TeV, D0 occupies an important position in the investigation of fundamental interactions at high energies. One of the main tools of the D0 experiment is the calorimeter, which is used to measure the energies and positions of electromagnetic particles and hadronic jets (remnants of constituents produced in $p\overline{p}$ collisions at large transverse momenta). To be of value, calorimeters must be calibrated using particles of known energy. Electrons and hadrons traversing calorimeter modules of the D0 detector were studied in a special test beam during the period between July 1991 and January 1992. A subset of the collected data is used here to improve the ability to measure the energy of hadronic jets. Using hadron data in the energy range 7.5-150 GeV, we optimize single-particle hadron energy resolution by varying the different layer weights (sampling weights) of the calorimeter that are used to parameterize the energy of incident particles. We find that through such optimizations we can achieve a 10% improvement in the energy resolution of single hadrons, and that this improvement is reduced to only about 3% when the weights are restricted to be energy-independent, position-independent, and fixed to the value that produces the best single-electron resolution. We observe that optimization causes the relative electronJhadron response to depart from unity, which suggests that weights based purely on energy loss through ionization (dE/dx), may be more suited for measuring jet energy, because they appear to produce a relative electronJhadron response ofthe calorimeter that is closer to unity. Using predictions from parton fragmentation, we then assemble electron and hadron events in the energy range between 2 GeV to 150 GeV into jets of energy between 15 GeV to 225 GeV. These jets should be very similar to jets observed at D0. We find that optimizing layer weights can improve the resolution of such simulated jets by a significant amount 00-15%) over that found using weights proportional to dE/dx. Our results can be used to develop techniques that could also improve the energy resolution ofjets observed in D0.