Simulating Outputs from the High-Luminosity Large Hadron Collider for Improved ATLAS Trigger Algorithm Testing

Abstract

The Large Hadron Collider is a proton-proton collider located at CERN near Geneva, Switzerland. It will undergo upgrades starting in 2026 to increase the amount of particle collisions produced by tenfold. This surge in data requires a more effective triggering system within the ATLAS detector located on the collider, so that only interesting collisions are stored for further analysis. This system receives full granularity detector data at 40 MHz, which is then processed by a series of algorithms. Finally, a trigger decision is made to keep or discard each event. One of these algorithms reconstructs topological clusters of energy from individual calorimeter cells. To study the performance of this algorithm, realistic physics simulations are used. The sample analyzed in this paper is simulated productions of two Higgs bosons, a rare event that could provide more insight into the properties of the Higgs boson. By reformatting the data of the sample into detector readout, the algorithm can be improved to better recognize di-Higgs events against a background of less interesting particle interactions. This thesis will describe the reformatting process in detail, as well as performance tests used to validate the simulated data and study clustering algorithms.