T2K: why look slightly away from the beam axis?

Choose an angle to choose a useful spectrum

J-PARC sends a predominantly muon-neutrino beam across Japan to Super-Kamiokande, about 295 kilometers away. T2K places the detector roughly two and a half degrees away from the beam’s central direction. That small offset changes the energy distribution reaching the detector; it does not bend a neutral neutrino after production.

A pion decays into a muon and a neutrino. In its rest frame, the two-body decay fixes the neutrino energy. Transforming to the laboratory links that energy to the pion’s boost and the neutrino’s angle. Looking slightly sideways compresses contributions from a range of energetic parents into a narrower band.

The gain has a cost: moving off axis sacrifices flux and does not produce a perfectly monochromatic source. Kaon decays, the parent distribution and the finite decay region still matter. The useful question is whether the resulting spectrum improves a particular oscillation measurement enough to justify that trade.

Sources: [1] · [2]

Match the window to a propagation question

The T2K spectrum peaks near six-tenths of a GeV, close to the first atmospheric-scale oscillation maximum for this baseline. A muon-flavor loss and an electron-flavor appearance can then be examined in the same deliberately chosen travel window.

Suppressing the higher-energy tail also helps control backgrounds and complicated interaction topologies. A narrower incident spectrum does not remove nuclear effects, however. The detector records outgoing particles, and translating their motion into incoming energy depends on the interaction and the nuclear target.

This distinguishes source design from event reconstruction. One shapes the distribution before the neutrinos arrive; the other estimates energy from each selected interaction afterward. A favorable source makes the inference easier, but an ideal source curve cannot replace a calibrated response.

Sources: [1] · [3] · [4]

Find a new flavor above an existing background

The source is mostly muon flavor, not exclusively muon flavor. Electron neutrinos already present in the beam can produce signal-like interactions. Neutral-current events can also imitate an electron ring when photons from a neutral pion are incompletely recognized.

T2K therefore combines beam predictions, near-detector measurements and far-detector event selection. The published 2014 appearance paper, submitted in November 2013, reported 28 selected electron-neutrino events against 4.92 expected background events, with a quoted background uncertainty of 0.55 and a 7.3-standard-deviation result.

Those figures describe that paper’s sample and analysis. They should not be mixed with the smaller 2013 evidence paper or the July 2013 announcement’s different numbers and significance. A publication trail records the evolution of an analysis, rather than one timeless count.

Sources: [3] · [5]

A nearby instrument supplies a constrained reference

The near detector measures a high-statistics interaction sample before the long journey. Beam direction monitors and tracking instruments supply complementary information. They constrain a joint description of the flux and interactions that will be used to predict the far record.

Near and far detectors do not automatically observe interchangeable events. Different targets, acceptance and selected final states can leave uncertainties that survive the extrapolation. Even a shared target cannot cancel a response error that changes with energy or topology.

The practical objective is to test how these surviving uncertainties affect the appearance inference. That is why cross-section measurements and detector calibrations belong inside the oscillation story. They determine whether a source designed for precision can deliver precision in the final fit.

Sources: [1] · [6]

Appearance opens a comparison with antimatter

Once electron-flavor appearance is established, neutrino- and antineutrino-dominated beam runs can probe how the probabilities differ. Horn polarity selects charged parents, so each run retains contamination and its own interaction rates. Equal detector counts would not mean equal probabilities.

The comparison also depends on matter along the route and on other mixing parameters. Reactors help constrain the electron-sector mixing angle, while disappearance supplies complementary information. The CP phase is inferred by fitting these ingredients together, not by attaching a label to one electron-like event.

T2K’s off-axis choice is consequently a design answer to a scientific question: concentrate events where the selected journey is informative. The next articles compare a longer American baseline and explain which asymmetries can support a claim about CP symmetry.

Sources: [1] · [4] · [6]

Try it in the Lab

Primary sources & revision

1. T2K Collaboration · The T2K Experiment (2011)
2. Adamson et al. · The NuMI Neutrino Beam (2015/2016)
3. T2K Collaboration · Observation of Electron Neutrino Appearance (2014)
4. Particle Data Group · Neutrino Masses, Mixing, and Oscillations (2024)
5. T2K Collaboration · Evidence of Electron Neutrino Appearance (2013)
6. T2K & NOvA · Joint neutrino oscillation analysis (22 October 2025)

First published 2026-10-11; last revised 2026-10-11. Original explanatory text and diagrams by Sandbox Physics. Illustrations are schematic; no experimental event records are reproduced here. This is an independent educational publication, not an official Nobel or experiment collaboration publication.

Continue the story