First bayesian neutrino oscillation sensitivities of dune
Author(s)
Warsame, Liban
Type
Thesis
Abstract
The Deep Underground Neutrino Experiment (DUNE) is a next-generation long-baseline neutrino oscillation experiment which aims to make world-leading precision measurements. DUNE will produce an intense beam of neutrinos at the Fermi National Accelerator Laboratory (Fermilab) directed at the Sandford Underground Research Facility (SURF) in South Dakota. Neutrino interactions will be measured by a near detector facility at Fermilab and a far detector facility at SURF. As a consequence of neutrino oscillation, DUNE will observe the disappearance of muon neutrinos along with the appearance of electron neutrinos. The observation of these neutrino oscillation channels provides DUNE with sensitivity to neutrino oscillation parameters.
To ensure that the experiment will achieve its physics goals, sensitivity studies are performed based on predictions of the neutrino event distributions that DUNE will observe. The results of these studies will also motivate and justify design choices that have an impact on sensitivity. In addition, DUNE will measure an order of magnitude more neutrino events than current long-baseline experiments. As a result, DUNE's sensitivity will be limited by systematic uncertainty, rather than statistical uncertainty. It is therefore imperative that these are modelled correctly in these studies to ensure accurately predicted sensitivities. Performing a Bayesian analysis allows for a more efficient method of studying the behaviour of systematic uncertainty parameters.
This thesis presents the first Bayesian neutrino oscillation sensitivities of DUNE. The Markov Chain Monte Carlo methods and Bayesian techniques used in this analysis will be described, as well as the implementation of systematic uncertainties which surpass that of previous DUNE sensitivity studies. The predicted sensitivities to the oscillation parameters will be shown, as well as the impact of systematic uncertainties. The deficiencies in the current systematic model are also discussed, with a demonstration of the level of complexity required in future analyses.
To ensure that the experiment will achieve its physics goals, sensitivity studies are performed based on predictions of the neutrino event distributions that DUNE will observe. The results of these studies will also motivate and justify design choices that have an impact on sensitivity. In addition, DUNE will measure an order of magnitude more neutrino events than current long-baseline experiments. As a result, DUNE's sensitivity will be limited by systematic uncertainty, rather than statistical uncertainty. It is therefore imperative that these are modelled correctly in these studies to ensure accurately predicted sensitivities. Performing a Bayesian analysis allows for a more efficient method of studying the behaviour of systematic uncertainty parameters.
This thesis presents the first Bayesian neutrino oscillation sensitivities of DUNE. The Markov Chain Monte Carlo methods and Bayesian techniques used in this analysis will be described, as well as the implementation of systematic uncertainties which surpass that of previous DUNE sensitivity studies. The predicted sensitivities to the oscillation parameters will be shown, as well as the impact of systematic uncertainties. The deficiencies in the current systematic model are also discussed, with a demonstration of the level of complexity required in future analyses.
Version
Open Access
Date Issued
2024-06-28
Date Awarded
01/03/2025
License URL
Advisor
Dunne, Patrick
Tapper, Alex
Publisher Department
Department of Physics
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)