Neutrino oscillation analysis using T2K data with additional single ring NCπ⁰ samples
File(s)
Author(s)
Martin, Daniel
Type
Thesis
Abstract
This thesis presents work done as part of the T2K and SK collaborations to perform an in-situ
reconstruction of PMT locations in Super-K and the addition of single ring and two ring NCπ0 event
samples to the T2K oscillation analysis. Photos taken by drone in Super-K were segmented using
machine learning algorithms to identify and reconstruct PMT positions. Feature recognition of PMTs
is shown to be viable, though it proved difficult to uniquely identify PMTs for reconstruction.
Markov Chain Monte-Carlo techniques were used as part of the MaCh3 software framework to per-
form an oscillation analysis with new NCπ⁰ samples using run 1-10 T2K data representing 19.664×10^20
POT in FHC and 16.346×10^20 POT in RHC. The dominant background to single ring electron-like ν_e
appearance samples at Super-K is NCπ⁰ events. The new NCπ0 samples aim to constrain the back-
ground such that additional oscillated ν_e events can be exploited for increased sensitivity in oscillation
fits. Super-K detector systematic uncertainties were updated to include the NCπ⁰ samples with a new
atmospheric neutrino fit, resulting in larger systematic uncertainties than previous analyses.
An oscillation fit with new NCπ⁰ samples and Super-K detector systematics gives credible intervals
of: sin²(θ_13) =[0.0185, 0.0417], sin²(θ_23) =[0.44, 0.58], ∆m²_23 =NO:[2.4, 2.6] IO:[-2.6, -2.4], and δ_CP
=[-π, -0.251] [2.551, π] to 2σ confidence. With the application of an additional sin²(θ_13) constraint
from reactor neutrino experiments, the intervals are: sin²(θ_23) =[0.45, 0.58], ∆m²_
23 =NO: [2.4, 2.6] IO:
[-2.6, -2.5], and δ_CP =[−π, -0.209] to 2σ confidence. Asimov fits show an additional 3.5% sensitivity
to δ_CP with the NCπ⁰ samples added. The credible intervals are in agreement with previous T2K
analyses and no further constraint to δ_CP is found. A slightly stronger preference for normal mass
ordering, upper sin²(θ_23) octant is observed compared to T2K with 62.1% posterior probability.
reconstruction of PMT locations in Super-K and the addition of single ring and two ring NCπ0 event
samples to the T2K oscillation analysis. Photos taken by drone in Super-K were segmented using
machine learning algorithms to identify and reconstruct PMT positions. Feature recognition of PMTs
is shown to be viable, though it proved difficult to uniquely identify PMTs for reconstruction.
Markov Chain Monte-Carlo techniques were used as part of the MaCh3 software framework to per-
form an oscillation analysis with new NCπ⁰ samples using run 1-10 T2K data representing 19.664×10^20
POT in FHC and 16.346×10^20 POT in RHC. The dominant background to single ring electron-like ν_e
appearance samples at Super-K is NCπ⁰ events. The new NCπ0 samples aim to constrain the back-
ground such that additional oscillated ν_e events can be exploited for increased sensitivity in oscillation
fits. Super-K detector systematic uncertainties were updated to include the NCπ⁰ samples with a new
atmospheric neutrino fit, resulting in larger systematic uncertainties than previous analyses.
An oscillation fit with new NCπ⁰ samples and Super-K detector systematics gives credible intervals
of: sin²(θ_13) =[0.0185, 0.0417], sin²(θ_23) =[0.44, 0.58], ∆m²_23 =NO:[2.4, 2.6] IO:[-2.6, -2.4], and δ_CP
=[-π, -0.251] [2.551, π] to 2σ confidence. With the application of an additional sin²(θ_13) constraint
from reactor neutrino experiments, the intervals are: sin²(θ_23) =[0.45, 0.58], ∆m²_
23 =NO: [2.4, 2.6] IO:
[-2.6, -2.5], and δ_CP =[−π, -0.209] to 2σ confidence. Asimov fits show an additional 3.5% sensitivity
to δ_CP with the NCπ⁰ samples added. The credible intervals are in agreement with previous T2K
analyses and no further constraint to δ_CP is found. A slightly stronger preference for normal mass
ordering, upper sin²(θ_23) octant is observed compared to T2K with 62.1% posterior probability.
Version
Open Access
Date Issued
2024-02
Date Awarded
2024-05
Copyright Statement
Creative Commons Attribution NonCommercial Licence
License URL
Advisor
Scott, Mark
Publisher Department
Physics
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)