学术文献库

SiD is a robust, silicon-based detector proposed for the International Linear Collider (ILC). SiD employs a sampling silicon-tungsten electromagnetic calorimeter (ECal) to accurately measure the energies of electrons, positrons, and photons produced in collisions, and to contribute to jet energy measurements through the particle flow technique. Due to the nature of the detector and its design constraints, a portion of the electron, positron, and photon energy exits the ECal undetected. Here, we establish a methodology for estimating the exiting energy and correctly determining the energies of electrons (and photons) in the ECal by analyzing patterns in the total energy deposition in each layer using neural networks. We studied a reduced calorimeter design with fewer layers (16 thin layers, 8 thick layers) than the proposed SiD design (20 thin layers, 10 thick layers) to evaluate if the resulting energy leakage could be determined and corrected. We evaluated the effectiveness of the correction on various electron energies and incidence angles in a Geant4 simple stack as well as in a modified version of the full DD4hep SiD model. The correction methodology showed significant improvement in measurement accuracy and resolution over uncorrected events, especially at high energies and shallow angles. These results provide a basis for effective energy correction in the SiD ECal and suggest that precise energy measurements with a smaller, less expensive ECal are viable.