Cosmic-ray all-electron spectrum with MAGIC
School
Ludwigs-Maximilians-Universität München
External url: https://edoc.ub.uni-muenchen.de/32865/
Abstract:
In this thesis, two methods are introduced for estimating the background cosmic-ray hadron events in order to further extract the electron and positron events based on Random Forest (RF) algorithm: the RF-Fit method, which is a template fit method, and the Two- Step RF method, which is a hard cut method based on two steps of RF training. The RF-Fit method is a commonly used technique for analyzing the cosmic-ray electron and positron spectrum with IACTs, which has been validated by instruments like H.E.S.S. and VERITAS. To adapt this method for use with MAGIC, I designed a Monte-Carlo (MC) tracking simulation method to ensure that the simulated background template precisely matches the Field of View (FoV) of the observation data. The Two-Step RF method is a novel approach that uses RF to accurately train between signal events and signal-like background events. By applying a tight cut of a few percent for the electron survival rate, the background events can be reduced to approximately 20%. After thorough evaluations of the systematics, both methods yield consistent reconstructed cosmic-ray electron and positron spectra in the energy range between 300 GeV and 6 TeV. The spectra can be described by a broken power-law and confirm the presence of an energy break around 900 GeV from MAGIC data for the first time, which is consistent with previous measurements. Afterwards, the contribution to the cosmic-ray electron and positron spectrum from Supernova Remnant (SNR) and pulsar models are compared with MAGIC data. The model preferred by MAGIC suggests that a Monogem pulsar is a source of the broken power law spec- trum, where the pulsar appears above the background but experiences suppression at high energies.