Sensitivity studies on tau neutrino appearance withKM3NeT/ORCA using Deep Learning Techniques
In the last few decades, it has been experimentally verified that neutrinos can change their flavor while propagating through space by so-called neutrino oscillations. The oscillation probabilities of neutrinos to oscillate from one flavor into another flavor are described by the neutrino mixing mat...
| Main Author: | |
|---|---|
| Format: | Doctoral or Postdoctoral Thesis |
| Language: | English |
| Published: |
2020
|
| Subjects: | |
| Online Access: | https://opus4.kobv.de/opus4-fau/frontdoor/index/index/docId/14214 https://nbn-resolving.org/urn:nbn:de:bvb:29-opus4-142148 https://opus4.kobv.de/opus4-fau/files/14214/phd_thesis_michael_moser_v2.pdf |
| Summary: | In the last few decades, it has been experimentally verified that neutrinos can change their flavor while propagating through space by so-called neutrino oscillations. The oscillation probabilities of neutrinos to oscillate from one flavor into another flavor are described by the neutrino mixing matrix. An important open question in particle physics is if the neutrino mixing matrix is unitary or not. Currently, the uncertainties on several matrix elements are too large in order to draw significant conclusions on the unitarity of the matrix. This is mostly due to the low experimental statistics in the tau neutrino sector. KM3NeT/ORCA is a water Cherenkov neutrino detector under construction in the Mediterranean Sea with several megatons of instrumented volume. Its main objective is the determination of the neutrino mass ordering. However, it will also observe about 4000 tau neutrino events per year, which will significantly improve the available tau neutrino statistics. In KM3NeT/ORCA, tau neutrinos will be identified by observing a statistical excess of charged-current-induced, cascade-like events with respect to the atmospheric electron neutrino expectation. The reconstruction of the low-level detector data consists of several stages. First, the detector background consisting of atmospheric muons and randomly correlated noise by 40K decays in seawater and by bioluminescence needs to be distinguished from the expected neutrino signals by using a classification algorithm. After this, track-like (charged-current muon neutrino) events are distinguished from cascade-like (charged-current electron neutrino, neutral-current electron neutrino) events based on another classifier. At last, neutrino properties like the energy and the direction of the neutrinos need to be reconstructed. Until now, maximum likelihood-based reconstruction algorithms accompanied by shallow machine learning techniques like Random Forests have been employed in the standard KM3NeT/ORCA reconstruction pipeline to tackle all of these tasks. A ... |
|---|