Detection of the Crab Nebula with the first TAIGA IACT

• Main Author: Blank, Michael
• Other Authors: Horns, Dieter
• Format: Doctoral or Postdoctoral Thesis
• Language: English
• Published: Staats- und Universitätsbibliothek Hamburg Carl von Ossietzky 2023
• Subjects: TAIGA (Experiment); Imaging Air Cherenkov Telescope (IACT); HiSCORE (Experiment); 530: Physik; 33.05: Experimentalphysik; 39.11: Astronomische Beobachtung; Observatorien; Planetarien; 39.12: Astronomische Instrumente; Gammaastronomie; Hochenergieastronomie; Luftschauer; Kosmische Strahlung; Čerenkov-Strahlung; Maschinelles Lernen; ddc:530: taiga
• Online Access: http://nbn-resolving.de/urn:nbn:de:gbv:18-ediss-111548; https://ediss.sub.uni-hamburg.de/handle/ediss/10436

The recent observation of gamma rays close to PeV energies has opened up a new, yet unexplored window to the sky. To identify the sources capable of producing such highly energetic particles and to understand the underlying acceleration mechanism, continued observations at energies above 100 TeV are a promising field of study. Instruments designed to observe at these energies require high sensitivity. The TAIGA experiment aims to implement a hybrid detection technique of cosmic and gamma rays at TeV to PeV energies. It combines the Cherenkov light sampling array HiSCORE with Imaging Air Cherenkov Telescopes (IACTs) and seeks to introduce a new reconstruction technique that complements the angular and shower core resolution of HiSCORE with the gamma hadron separation power of the imaging telescopes. In this work, a simulation and analysis chain of the first TAIGA IACT starting from raw data and concluding with a high-level spectral reconstruction of Crab Nebula data of only the first IACT, taken between October 2019 and December 2020 is presented. A good agreement of higher-level image parameters with simulation is demonstrated. Using the image parameters, origin reconstruction, energy estimation, and background suppression were implemented with Random Forests, a commonly used machine learning technique. Overall, an angular resolution of <0.2° between 10 TeV and 100 TeV degrading towards higher and lower energies was achieved. The energy resolution is <25 % and the bias is <5 % between 10 TeV and 100 TeV. For the background suppression with the Random Forest, a quality factor of 4 for a Crab-like spectrum was achieved. Above 30 TeV, a quality factor of 6 was reached. Together with the instrument response functions from the Random Forest analysis, calibrated, cleaned, reconstructed, and background suppressed events were used in the higher-level analysis with Gammapy. 122 h of good-weather-selected Crab Nebula data taken in wobble mode resulted in 204.6 excess events at a significance of 9.7σ for an ...