Investigating Calibration Procedures for the ATLAS Trigger System

The ATLAS calorimeter is a non-compensating calorimeter that was optimised for measuring electrons and photons. This means that for strongly interacting particles (hadrons), their energy is only partially measured. For example, typically only 70% of a charged pion’s energy is measured in the calorimeter. The rest of the energy is invisible, and is released in nuclear reactions, in low energy neutrons, and as heat. Proton-proton collisions induce hard scattering interactions, producing quarks and gluons. These particles leave jets, which can then be measured by the detector. The energy of jets is also only partially measured. The ATLAS detector has a two-level trigger system. It consists of a Level 1 (L1) electronics-based trigger system, and the high-level trigger (HLT) system, which is a software-based system. L1 and HLT jets refer to jets that are based off of their respective trigger systems. The ATLAS hadron calibration procedure uses Monte Carlo simulations to correct for this inefficieny in measuring the energy on average. In addition to simulating the L1 and HLT systems, the Monte Carlo simulations also contain a truth jets; these correspond to the true energy and momentum of the particles. In order to improve the energy resolution, variables indicating the type of dominant interactions for a given jet can be used. One of these variables is the electromagnetic fraction defined as the energy deposited in the calorimeter divided by the total energy of the particle. If we only consider pions (the lightest hadrons), a jet contains typically 30% of neutral, negatively or positively charged pions. Unlike the charged pions, the neutral pions decay dominantly into two photons and therefore their energy is fully measured in the calorimeter. Consequently, jets with a high neutral pion content have a high response and a large energy deposit in the electromagnetic calorimeter which contains the photons from the pion decay. The jet response r is defined as the ratio between the pT of two types of jets. For example, the HLT-truth response is the ratio of HLT jet pT to the truth jet pT . The jet resolution can be improved by studying r as a function of the EM fraction. The goal of this project is to investigate if these kinds of calibrations can be also used in the ATLAS trigger system. During the project, the main datasets used were Run 2 data and Monte Carlo simulation data. A Run 3 dataset was also considered, but was abandoned due to a lack of available statistics. All data analysis was done using RDataFrame.