Mass-squared differences
The differences between squared neutrino masses, which set how fast oscillations happen.
Δm²ij = mi² - mj². Only two are independent: the small 'solar' splitting Δm²21 and the ~33 times larger 'atmospheric' splitting Δm²31 (or Δm²32). Their ratio creates a hierarchy of oscillation lengths.
Current measured values of all neutrino parameters →
In plain words
Oscillations are a beat phenomenon. Two tuning forks with slightly different pitches produce a slow rise and fall in loudness whose rate depends only on the difference between the pitches, not on the pitches themselves. Neutrino oscillations work the same way: their frequency depends on the differences between the squared masses of the mass states. That is why oscillation experiments can say a great deal about mass differences and almost nothing about how heavy any single neutrino is.
In detail
With three mass states there are three differences, Δm²21, Δm²31 and Δm²32, but they obey Δm²31 = Δm²32 + Δm²21, so only two are independent. One is small and was first seen through solar neutrinos, the other is about 33 times larger and was first seen through atmospheric neutrinos:
- Δm²21 ≈ 7.5 × 10⁻⁵ eV², the solar splitting, sign known to be positive.
- Δm²31 ≈ ±2.5 × 10⁻³ eV², the atmospheric splitting, sign unknown.
The unknown sign of the large splitting is the question of the mass ordering. Because the two splittings differ so much, most experiments are sensitive to one at a time, which justifies the two-flavour approximation.
In numbers
The splittings set lower limits on the masses. In normal ordering the heaviest state has at least √(2.5 × 10⁻³ eV²) ≈ 0.05 eV and the middle one at least √(7.5 × 10⁻⁵ eV²) ≈ 0.009 eV. Hence at least two neutrinos are massive. Direct experiments such as KATRIN and cosmological bounds on the sum of neutrino masses provide the missing upper limits.
Why it matters for neutrino physics
The two splittings fix the two natural length scales of oscillation and therefore the placement of experiments. Accelerator, atmospheric and short-baseline reactor experiments target the atmospheric scale; long-baseline reactor and solar experiments target the solar scale. Their precise values, now known at the percent level, also feed predictions for neutrinoless double beta decay and cosmology.
Frequently asked questions
Why squared masses and not masses?
Relativistic energy is E ≈ p + m²/2p, so the phase difference between mass states depends on m², and the observable is the difference mi² − mj².
Could there be more mass splittings?
Only if more neutrino states exist. Some short-baseline anomalies have been interpreted as a sterile neutrino with Δm² around 1 eV², but this is not established.
Original papers
- DiscoveryY. Fukuda et al. (Super-Kamiokande Collaboration), Evidence for Oscillation of Atmospheric Neutrinos, Phys. Rev. Lett. 81, 1562 (1998)DOI 10.1103/PhysRevLett.81.1562
- ReviewI. Esteban, M. C. Gonzalez-Garcia, M. Maltoni, T. Schwetz and A. Zhou, The fate of hints: updated global analysis of three-flavor neutrino oscillations, JHEP 09, 178 (2020)DOI 10.1007/JHEP09(2020)178
Go deeper
- Solar mass splitting Δm²21The small mass-squared difference between ν2 and ν1.
- Atmospheric mass splitting Δm²31The large mass-squared difference involving ν3.
Related terms
Mentioned in 75 pages
Lessons 5
FAQ 4
Milestones 6
- 1995 · LSND reports an unexpected oscillation signal
- 1998 · CHOOZ sees no reactor antineutrino disappearance
- 1999 · K2K starts the first long-baseline experiment
- 2004 · KamLAND observes the oscillation spectrum
- 2004 · Super-K sees the oscillation dip in L/E
- 2007 · MiniBooNE tests LSND and finds a low-energy excess
People 5
Terms 53
- Appearance vs disappearance
- Baseline
- Beryllium-7 neutrinos
- BEST (Baksan Experiment on Sterile Transitions)
- BNL E776
- CDHS (WA1)
- CERN West Area Neutrino Facility (WANF)
- Charged lepton flavour violation
45 more
- CHOOZ
- CHORUS (WA95)
- Disappearance measurement
- Discovery of neutrino oscillations
- Effective Electron-neutrino mass
- Energy-momentum relation
- eV and Sub-eV (mass scale)
- Gallium anomaly
- GeV scale
- Global fit
- Gösgen reactor experiment
- ICARUS
- ILL Grenoble reactor experiment
- K2K
- L/E
- L/E (baseline over energy)
- Large mixing angle (LMA) MSW solution
- Look-elsewhere effect
- LSND
- MACRO
- Matter effects
- Matter potential
- MINOS / MINOS+
- Mixing angle θ13
- MSW effect
- MSW resonance condition
- Neutrino energy reconstruction
- Neutrino mass ordering
- Neutrino oscillation physics
- NOMAD (WA96)
- Nuclear models for neutrino scattering
- Palo Verde
- Pseudo-Dirac neutrino
- Rate vs shape analysis
- Reactor antineutrino anomaly
- RENO
- Short-baseline & Sterile-neutrino searches
- Short-baseline anomalies
- Solar mass splitting Δm²21
- Soudan 2
- Two-flavour approximation
- Units and relativistic kinematics
- Upper limit / exclusion
- Vacuum oscillations
- Zenith angle
Last updated: · Editorial standard
Cite this page
Text
Neutrino Atlas (2026). “Mass-squared differences”. Retrieved 22 November 2026, from https://neutrino-atlas.com/terms/mass-squared-differences/BibTeX
@misc{neutrinoatlas2026msmasssquareddifferences,
title = {Mass-squared differences},
author = {{Neutrino Atlas}},
year = {2026},
url = {https://neutrino-atlas.com/terms/mass-squared-differences/},
note = {Accessed 2026-11-22},
language = {english}
}