Emily Marie Duffield (Lawrence Berkeley National Lab. (US))
Reconstruction of charged particles in dense environments is paramount for a number of applications including jet energy and mass calibration, jet flavour tagging, and the reconstruction of hadronic decays from high momentum taus. With the center of mass energy increase for Run 2 of the LHC, all of these are of special importance, for example in the reconstruction of high transverse momentum objects from potential heavy resonances. In these environments, it is not uncommon for multiple particles to deposit energy in the same or nearby pixels in the ATLAS pixel detector, which results in a single merged cluster during track reconstruction. Methods to identify such merged measurements are implemented in the ATLAS software, but a residual inefficiency in reconstructing nearby tracks remains, resulting in lost tracks. In this poster, a fully data driven method to determine this remaining inefficiency is presented. The method uses the ionization loss per unit length (dE/dx) in the silicon sensors to determine the fraction of lost tracks. Results using 13 TeV proton-proton collision data collected at the LHC in 2015 are shown and compared to Monte Carlo simulation. The fraction of lost tracks in data is determined to range from 1.0% to 3.6% for 200-1600 GeV jets with simulation agreeing within 40%.
ATLAS Collaboration (CERN)