T2Kvs.NOvA
T2K vs. NOvA: two long-baseline experiments compared
In short
Both send off-axis muon-neutrino beams to distant detectors to measure νe appearance, δCP and the mass ordering. T2K uses 295 km and Super-Kamiokande; NOvA uses 810 km and a scintillator detector, with stronger matter effects. Their first joint analysis appeared in Nature in October 2025.
| Criterion | T2K | NOvA |
|---|---|---|
| Beam source | J-PARC, Tokai, Japan | NuMI beam, Fermilab, USA |
| Far detector | Super-Kamiokande, 50,000 t water | 14,000 t segmented liquid scintillator |
| Baseline | 295 km | 810 km |
| Off-axis angle | 2.5° | About 0.8° (14 mrad) |
| Peak energy | About 0.6 GeV | About 2 GeV |
| Matter effect | Small | About three times larger |
| Near detector | ND280 at 280 m, different technology | Functionally identical to far detector |
| Beam data since | 2010 | 2014 |
| Successor | Hyper-Kamiokande, around 2028 | DUNE, early 2030s |
Same strategy, different scales
T2K and NOvA are accelerator long-baseline experiments. Each produces a muon-neutrino or muon-antineutrino beam, measures it in a near detector and looks for νμ disappearance and νe appearance in a far detector. Both place the far detector slightly off the beam axis. This off-axis trick produces a narrow energy spectrum tuned to the first oscillation maximum. T2K sends its beam 295 km from J-PARC to Super-Kamiokande, 1,000 m underground, peaking near 0.6 GeV. NOvA sends the NuMI beam 810 km from Fermilab to Ash River, Minnesota, peaking near 2 GeV. Its 14,000-tonne far detector is built from PVC cells filled with liquid scintillator and sits near the surface.
Why the differences matter
The νe appearance probability depends on θ₁₃, θ₂₃, δCP and the mass ordering. Matter effects grow with energy and baseline, so they are about three times larger in NOvA than in T2K. In T2K, the ordering and δCP shift the appearance rates in nearly the same way. In NOvA they can pull in opposite directions, so a single experiment cannot always separate the two. Combining a baseline with weak matter effects and one with stronger matter effects breaks part of this degeneracy. The experiments also differ in nuclear targets, detector technology and cross-section models.
The joint analysis
Taken separately, T2K favoured δCP near −π/2 while NOvA's data under normal ordering allowed a wide range of values, a mild tension. Their first joint analysis, published in Nature in October 2025, combined about a decade of T2K data with six years of NOvA data. It measured |Δm²₃₂| to better than 2 percent and found no strong preference for either ordering. If the ordering is inverted, the data provide evidence for CP violation in the lepton sector, with CP-conserving values disfavoured at more than 3σ. Hyper-Kamiokande and DUNE will continue the two lines.
Read on
- Lesson: Neutrino beams and long-baseline experiments →
- Lesson: Mixing and CP violation →
- FAQ: What is CP violation in neutrinos? →
- FAQ: What is the neutrino mass ordering? →
Related terms
- T2K
- NOvA
- Accelerator Long-baseline experiments
- Off-axis beam
- CP-Violating phase δCP
- Oscillation parameter degeneracies
Last updated: · Editorial standard
Cite this page
Text
Neutrino Atlas (2026). “T2K vs. NOvA: two long-baseline experiments compared”. Retrieved 22 November 2026, from https://neutrino-atlas.com/comparisons/t2k-vs-nova/BibTeX
@misc{neutrinoatlas2026comparisonst2kvsnova,
title = {T2K vs. NOvA: two long-baseline experiments compared},
author = {{Neutrino Atlas}},
year = {2026},
url = {https://neutrino-atlas.com/comparisons/t2k-vs-nova/},
note = {Accessed 2026-11-22},
language = {english}
}