On June 5, 1998, at the Neutrino ‘98 conference in Takayama, Japan, Takaaki Kajita of the University of Tokyo walked to the podium and presented a result that broke the Standard Model.
The data came from Super-Kamiokande — a 50,000-ton water Cherenkov detector buried a kilometer beneath the Japanese Alps — and it showed something specific and undeniable: muon neutrinos produced in Earth’s atmosphere were vanishing in a way that depended on how far they had traveled before hitting the detector. The most natural explanation was that they were oscillating into another flavor of neutrino, and oscillation requires that neutrinos have mass.
The Standard Model, as written in the 1970s, had assumed they were massless. Super-Kamiokande’s announcement meant the model was incomplete. Seventeen years later, Kajita shared the Nobel Prize in Physics with SNO’s Art McDonald for that result.
The atmospheric neutrino anomaly
The story did not start in 1998. For more than a decade, several earlier experiments — Kamiokande, IMB, Soudan-2 — had been quietly recording a strange feature in their atmospheric neutrino data. They were seeing fewer muon neutrinos than the calculation predicted, and the deficit seemed to depend on the direction the neutrinos came from.
Atmospheric neutrinos are made when cosmic rays hit Earth’s upper atmosphere. The collisions produce pions and kaons, which decay into muons and muon neutrinos. The muons themselves decay into electrons plus more neutrinos. The end result is a well-understood flux of atmospheric neutrinos, dominantly muon-flavored, raining down on the Earth from all directions.
The “all directions” part is important. A detector deep underground sees atmospheric neutrinos coming down from the sky above (having traveled only a few kilometers through the atmosphere) and also coming up from below (having traveled through the entire Earth, about 13,000 kilometers, after being produced on the opposite side of the planet). At neutrino energies of a few GeV, the cross-section is so small that the Earth is essentially transparent — so the rate of upward-going muon neutrinos should match the rate of downward-going ones.
The atmospheric anomaly was that the upward-going muon neutrinos were missing in larger numbers than the downward ones.
What Super-K saw
Super-Kamiokande, which started taking data in April 1996, was much larger than the previous generation of experiments. By 1998 it had accumulated about 535 days of running and many thousands of atmospheric neutrino events.
The analysis the team presented in Takayama plotted the ratio of measured to expected muon-neutrino rate as a function of the zenith angle — the angle from straight up. Downward-going neutrinos (zenith angle near 0°) gave roughly the expected rate. Upward-going neutrinos (zenith angle near 180°) gave only about half. The transition between the two was smooth and well-fit by the curve predicted by neutrino oscillation.
In a flat 5σ result — high enough that statistical fluctuation was effectively ruled out — Super-Kamiokande announced that atmospheric muon neutrinos oscillate.
The interpretation was that muon neutrinos transformed primarily into tau neutrinos over the long baseline through the Earth. Tau neutrinos at those energies do not produce a clean detection signature in Super-K, so they essentially “disappeared” from the count.
Why this required neutrino mass
Neutrino oscillation, mathematically, requires two things.
Mixing between flavor and mass eigenstates. The particle states that participate in the weak interaction (electron, muon, tau flavors) are not the same as the states with definite mass. Each flavor state is a quantum-mechanical superposition of two or three mass states.
A difference between the squared masses. As a neutrino travels, the different mass components of its superposition develop phase factors that oscillate at different rates. The interference between them produces the flavor-changing signature. If all the masses were equal — and in particular, if all of them were zero — there would be no oscillation.
So the moment Super-Kamiokande’s curve fit oscillation, the result implied at least one neutrino species has mass. It did not say which one, or how much, or whether the others were also massive. But it forced an addition to the Standard Model that had to be reckoned with.
The broader consequences
Super-K’s 1998 announcement was the first major result in what is now called the era of precision neutrino physics. Several immediate consequences followed.
The Standard Model needed extending. Some mechanism — Dirac mass, Majorana mass via the see-saw mechanism, or something more exotic — had to be added to explain why neutrinos are massive but very light.
The solar neutrino problem gained a natural explanation. If atmospheric neutrinos could oscillate, then solar neutrinos probably could too. This insight was confirmed by the SNO experiment in 2001.
A measurement program emerged. Once oscillation was established, the next questions naturally followed: what are the mixing angles, what are the mass-squared differences, is there CP violation, what is the mass ordering? Each of these has been the subject of major experimental campaigns since.
Cosmology incorporated massive neutrinos. The radiation density of the early universe, the formation of cosmic structure, and the cosmic microwave background all had to be re-analyzed with non-zero neutrino mass. This is now a routine part of cosmological model-fitting.
How the detector itself works
Super-Kamiokande consists of a cylindrical stainless-steel tank — 39 meters in diameter and 41 meters tall — filled with 50,000 tons of ultra-pure water. Lining the inner wall of the tank are 11,146 photomultiplier tubes, each 50 centimeters across, which collectively cover about 40% of the inner surface.
When a neutrino interacts in the water, it produces a charged particle — usually an electron (for electron-neutrino interactions) or a muon (for muon-neutrino interactions). The charged particle travels faster than light in water, emitting a cone of Cherenkov radiation behind it. The light hits the photomultipliers, which record the timing and intensity at each tube.
The pattern of light on the inner wall reveals the trajectory and energy of the particle. Electrons produce diffuse, fuzzy Cherenkov rings (because they scatter as they travel). Muons produce sharp, clean rings (because they travel almost straight). The shape of the ring tells the analysts whether the neutrino was electron-flavored or muon-flavored.
After 1998
Super-Kamiokande continued to run continuously, and was the far detector for the K2K and T2K experiments. In 2020 the tank was emptied for the first time in over twenty years to load gadolinium sulfate into the water — an upgrade that allows the detector to distinguish electron antineutrinos from electron neutrinos by detecting the neutron capture on gadolinium that follows inverse beta decay. The upgraded Super-K is now searching for the diffuse supernova neutrino background, the integrated relic flux from all supernovae throughout cosmic history.
The next-generation Hyper-Kamiokande detector, eight times the size of Super-K, is being built nearby. It will start operations in 2027 and will inherit Super-K’s role as one of the central instruments of neutrino physics.
The legacy of one summer day
The June 1998 announcement is one of the great moments in modern physics. A clean, statistically convincing result, presented in a single talk, that immediately overturned a textbook assumption. The Standard Model — beautifully successful at predicting almost everything — had a hole in it that nobody had been sure existed before.
That hole is now its own field of study. Neutrino oscillation, mass, Majorana versus Dirac character, CP violation — all of it traces directly back to Kajita’s talk in Takayama, and to the 535 days of patient running by the world’s largest water Cherenkov detector underneath a Japanese mountain.
For the parallel discovery in the solar sector, see SNO. For the experimental aftermath, see T2K and Hyper-Kamiokande. For the mechanism behind oscillation, see How neutrino oscillation works.
Further reading
Primary sources
- Fukuda et al. (Super-Kamiokande), “Evidence for oscillation of atmospheric neutrinos”, Phys. Rev. Lett. 81:1562 (1998) — the discovery paper
- Takaaki Kajita’s 2015 Nobel Lecture — Kajita’s own retelling of the discovery
Background and context
- Super-Kamiokande official site (ICRR, University of Tokyo) — detector specifications and ongoing physics
- Wikipedia: Super-Kamiokande
- Symmetry Magazine — “Twenty years of neutrino oscillation” — anniversary feature on the 1998 result
Frequently asked
What did Super-Kamiokande discover in 1998?
The team announced that muon neutrinos produced in cosmic-ray interactions in Earth's atmosphere disappear as they travel further before reaching the detector. The disappearance rate matched neutrino oscillation — the conversion of muon neutrinos into another flavor — and required neutrinos to have non-zero mass. It was the first direct evidence that neutrinos are massive particles.
Why did the 1998 result win a Nobel Prize?
Because the Standard Model of particle physics had assumed neutrinos were massless. Super-K's result, combined with SNO's later confirmation using solar neutrinos, proved this assumption wrong. Takaaki Kajita shared the 2015 Nobel Prize in Physics with SNO's Art McDonald for the discovery of neutrino oscillation and therefore of neutrino mass.
What are atmospheric neutrinos?
Atmospheric neutrinos are produced when cosmic rays — high-energy protons from space — collide with atoms in Earth's upper atmosphere. The collisions create showers of pions and muons, which decay into neutrinos. About two muon neutrinos are produced for every electron neutrino, providing a known reference flux that experiments can compare against.
What is neutrino oscillation?
Neutrino oscillation is the quantum-mechanical phenomenon in which a neutrino produced in one flavor — say a muon neutrino — has a probability of being detected later as a different flavor, like a tau neutrino. The oscillation only occurs if neutrinos have non-zero, non-degenerate masses, so its observation directly implies neutrino mass.
Where is Super-Kamiokande located?
Super-Kamiokande sits about 1,000 meters underground in the Mozumi mine near the village of Kamioka in Gifu Prefecture, Japan. The mountain above shields the detector from cosmic-ray backgrounds. The facility is operated by the University of Tokyo's Institute for Cosmic Ray Research.
Cite this article 5 formats
APA
Neutrino Times Editorial Team. (2025, October 4). Super-Kamiokande 1998: the announcement that gave neutrinos mass. Neutrino Times. https://neutrino-times.com/articles/super-kamiokande-1998-atmospheric-oscillation-discovery/
Chicago
Neutrino Times Editorial Team. "Super-Kamiokande 1998: the announcement that gave neutrinos mass." Neutrino Times, October 4, 2025. https://neutrino-times.com/articles/super-kamiokande-1998-atmospheric-oscillation-discovery/.
MLA
Neutrino Times Editorial Team. "Super-Kamiokande 1998: the announcement that gave neutrinos mass." Neutrino Times, 4 Oct. 2025, https://neutrino-times.com/articles/super-kamiokande-1998-atmospheric-oscillation-discovery/.
BibTeX
@misc{neutrino-times-super-kamiokande-1998-atmospheric-oscillation-discovery,
author = {Neutrino Times Editorial Team},
title = {Super-Kamiokande 1998: the announcement that gave neutrinos mass},
howpublished = {Neutrino Times},
year = {2025},
month = {oct},
url = {https://neutrino-times.com/articles/super-kamiokande-1998-atmospheric-oscillation-discovery/},
note = {Accessed: 2025-10-04}
} RIS
TY - GEN TI - Super-Kamiokande 1998: the announcement that gave neutrinos mass AU - Neutrino Times Editorial Team PY - 2025 DA - 2025-10-04 PB - Neutrino Times UR - https://neutrino-times.com/articles/super-kamiokande-1998-atmospheric-oscillation-discovery/ ER -