Results from the T2K experiment on neutrino mixing including a new far detector $μ$-like sample

TL;DR

The paper reports a three-flavor neutrino oscillation measurement with T2K using $3.6\times10^{21}$ POT at SK, featuring major upgrades to flux and interaction modeling and a new $\nu_\mu$ CC1$\pi^+$-like sample in the far detector. It introduces new near-detector samples (proton-tagged and photon-tagged) and an expanded far-detector muon-neutrino sample with tagged pions, along with NA61/SHINE replica-target constraints that reduce the $\nu_\mu$ flux uncertainty to below 4% up to 7 GeV. Oscillation analysis employs both MCMC and frequentist (FC) approaches, with a Reactor Constraint applied in some fits, and consistency checks via posterior-predictive p-values. In the normal mass ordering, the best-fit parameters are $\sin^2\theta_{23}=0.559^{+0.018}_{-0.078}$, $\Delta m^2_{32}=(2.506^{+0.039}_{-0.052})\times10^{-3}\text{ eV}^2$, and $\delta_{CP}=-2.18^{+1.22}_{-0.47}$, indicating a preference for normal ordering and the upper octant and excluding certain CP-conserving values at significant confidence levels; a data release of MCMC chains without the Reactor Constraint is provided. These results reflect a robust, model-bounded neutrino-oscillation analysis with improved control of interaction uncertainties, paving the way for future precision in long-baseline experiments.

Abstract

We have made improved measurements of three-flavor neutrino mixing with 19.7(16.3)$\times 10^{20}$ protons on target in (anti-)neutrino-enhanced beam modes. A new sample of muon-neutrino events with tagged pions has been added at the far detector, as well as new proton and photon-tagged samples at the near detector. Significant improvements have been made to the flux and neutrino interaction modeling. T2K data continue to prefer the normal mass ordering and upper octant of $\sin^2θ_{23}$ with a near-maximal value of the charge-parity violating phase with best-fit values in the normal ordering of $δ_{\scriptscriptstyle\mathrm{CP}}=-2.18\substack{+1.22 \\ -0.47}$, $\sin^2θ_{23}=0.559\substack{+0.018 \\ -0.078}$ and $Δm^2_{32}=(+2.506\substack{+0.039 \\ -0.052})\times 10^{-3}$ eV$^{2}$.

Figures (6)

• Figure 1: Data and model predictions before and after the ND280 fit for the $\nu$-mode FGD1 CC0$\pi$0p sample binned in $p_{\mu}$. The fit includes $p_{\mu} \leq 30$ GeV for all ND280 samples, but the range is truncated here for readability. The postfit prediction is broken down by interaction channel. The predicted event rate from $\bar{\nu}$ contamination in $\nu$-mode is neglected in the figure as it only contributes at the sub-percent level.
• Figure 2: The $E^\text{rec}_{\nu}$ distribution for the $\nu_{\mu}$CC1$\pi^+$$\nu$-mode SK sample shown for data and MC. The fit includes $\mu$-like ($e$-like) events with $E^\text{rec}_{\nu} \leq 30$ (1.25) GeV, but the range is truncated here for readability. The reference oscillation parameter values are described in the main text. The ND280 constraint has been applied to the MC prediction.
• Figure 3: Credible regions in the $\sin^2\theta_{13}$--$\delta_{\mathrm{CP}}$ plane produced with the MCMC analysis, shown with and without the RC applied, and overlaid with the RC constraint from Ref. PDG2021, for normal ordering.
• Figure 4: The change in the best fit $\chi^{2}$ observed in the frequentist analysis as a function of $\delta_{\mathrm{CP}}$ and the mass ordering. Shaded regions correspond to the FC-corrected confidence intervals.
• Figure 5: Frequentist confidence intervals are shown in the $\sin^2\theta_{23}$--$\Delta{}m^2_{32}$ plane, produced using the constant $\Delta\chi^{2}$ method with the RC applied. The $\Delta{}m^2_{32}$ contour is smeared to cover the SDS as described in the main text.