PMNS unitarity tests
Checking that the three known neutrino flavours fully account for mixing.
If the PMNS matrix is exactly unitary, rows and columns must each have squared moduli summing to 1. Deviations would indicate extra (sterile or heavy) neutrinos. Current tests reach the few-percent level for the electron row and are weaker for the μ and τ rows.
In plain words
The PMNS matrix translates between the three flavours and the three mass states. If there are exactly three light neutrinos and nothing else mixes with them, the matrix is like a perfect rotation: the probabilities of finding an electron neutrino in each of the three mass states must add up to exactly one. If some probability goes missing, it is leaking into states we have not seen.
In detail
Unitarity means U U† = 1. It gives two kinds of conditions: normalisations, such as |Ue1|² + |Ue2|² + |Ue3|² = 1 for each row and column, and closure relations between pairs of rows, such as Ue1 U*_μ1 + Ue2 U*_μ2 + Ue3 U*_μ3 = 0. In the standard three-flavour analysis unitarity is assumed; a test fits all nine elements more freely and checks whether the conditions hold.
Two broad kinds of new physics break it. Light sterile neutrinos that mix with the active ones produce extra oscillation frequencies. Heavy states, such as the heavy neutral leptons of seesaw models, are too heavy to be produced in oscillation experiments, so the visible 3 × 3 block is simply not unitary, which shows up as an overall deficit or a small flavour change at zero distance.
In numbers
- Parke and Ross-Lonergan (2016) found, from oscillation data alone at 3σ, that the electron and muon row normalisations deviate from one by at most about 7 %, while the other rows and columns are only known to within 20–40 %.
- The tau row is least constrained, because tau neutrinos are hard to produce and detect; only a few experiments such as OPERA and Super-Kamiokande have identified ντ appearance.
- If the extra states are much heavier than the electroweak scale, electroweak precision data and limits on lepton flavour violation constrain deviations far more strongly, roughly at the 10⁻³ level.
Why it matters for neutrino physics
Testing unitarity turns oscillation measurements into a search for new particles that does not rely on any specific model. Precise measurements of νe and ντ appearance at DUNE, Hyper-Kamiokande and JUNO, and future ντ samples, are expected to tighten the tests considerably.
Frequently asked questions
Does the number of light neutrinos from the Z boson already prove unitarity?
No. The Z width shows there are three light active neutrinos, but heavy states or light sterile states mixing with them can still make the 3 × 3 matrix slightly non-unitary.
Original papers
- TheoryS. Antusch, C. Biggio, E. Fernández-Martínez, M. B. Gavela and J. López-Pavón, Unitarity of the Leptonic Mixing Matrix, JHEP 10, 084 (2006)
- MeasurementS. Parke and M. Ross-Lonergan, Unitarity and the three flavor neutrino mixing matrix, Phys. Rev. D 93, 113009 (2016)DOI 10.1103/PhysRevD.93.113009
Related terms
Mentioned in 6 pages
Last updated: · Editorial standard
Cite this page
Text
Neutrino Atlas (2026). “PMNS unitarity tests”. Retrieved 22 November 2026, from https://neutrino-atlas.com/terms/pmns-unitarity/BibTeX
@misc{neutrinoatlas2026termspmnsunitarity,
title = {PMNS unitarity tests},
author = {{Neutrino Atlas}},
year = {2026},
url = {https://neutrino-atlas.com/terms/pmns-unitarity/},
note = {Accessed 2026-11-22},
language = {english}
}