Measurement of atmospheric neutrino oscillation parameters using convolutional neural networks with 9.3 years of data in IceCube DeepCore

TL;DR

This paper reports a precise measurement of atmospheric neutrino oscillation parameters using 9.3 years of IceCube DeepCore data, employing convolutional neural networks for improved event reconstruction and background suppression.

Contribution

The study introduces a convolutional neural network-based reconstruction method that enhances signal efficiency and background rejection in atmospheric neutrino analysis.

Findings

Measured m_{32} = 2.40^{+0.05}_{-0.04} imes 10^{-3} eV^2

Measured ^2_{23} = 0.54^{+0.04}_{-0.03}

Results are the most precise atmospheric neutrino oscillation measurements to date.

Abstract

The DeepCore sub-detector of the IceCube Neutrino Observatory provides access to neutrinos with energies above approximately 5 GeV. Data taken between 2012-2021 (3,387 days) are utilized for an atmospheric $\nu_\mu$ disappearance analysis that studied 150,257 neutrino-candidate events with reconstructed energies between 5-100 GeV. An advanced reconstruction based on a convolutional neural network is applied, providing increased signal efficiency and background suppression, resulting in a measurement with both significantly increased statistics compared to previous DeepCore oscillation results and high neutrino purity. For the normal neutrino mass ordering, the atmospheric neutrino oscillation parameters and their 1$\sigma$ errors are measured to be $\Delta$m$^2_{32}$ = $2.40\substack{+0.05 \\ -0.04} \times 10^{-3} \textrm{ eV}^2$ and sin$^2$$\theta_{23}$=$0.54\substack{+0.04 \\ -0.03}$. The results are the most precise to date using atmospheric neutrinos, and are compatible with measurements from other neutrino detectors including long-baseline accelerator experiments.