Cherenkov Water Tank

A buried ocean of ultrapure water, listening for the blue light that neutrinos leave behind.

Neutrinos almost never interact with anything, which makes detecting them a numbers game: the more water you have, the better your chances. When a neutrino does interact, it produces a charged secondary particle — usually an electron or muon — that moves faster than the speed of light in water (about 75% of the vacuum speed). The resulting Cherenkov flash is detected by arrays of photomultiplier tubes lining the tank walls. The ring's sharpness distinguishes particle types, its brightness encodes energy, and its timing encodes direction. Used the same way in Antarctic ice (IceCube), the technique scales to cubic-kilometer volumes.

Light travels at roughly 225,000 km/s in water, because the refractive index of water is about 1.33. A relativistic charged particle that exceeds this speed emits coherent Cherenkov radiation at a fixed half-angle given by cos θ = 1/(nβ), where n is the refractive index and β = v/c. In water this angle is approximately 41° for ultra-relativistic particles. Photomultiplier tubes recording the arrival times of photons at hundreds to thousands of points across the tank wall allow reconstruction of the cone axis — and hence the incoming neutrino direction — to within a few degrees. Electron-like particles scatter electromagnetically and produce diffuse, fuzzy rings; muons travel straighter and produce sharp rings. This difference in ring topology is what let Super-Kamiokande separate electron neutrino events from muon neutrino events without a magnetic field. At IceCube, where sensors sit in Antarctic ice rather than water, the geometry is cylindrical and the instrumented volume reaches one cubic kilometer, but the underlying Cherenkov principle is identical.

In plain terms

You know how a jet creates a sonic boom when it flies faster than sound? Cherenkov light is the same idea but for light. A charged particle zips through water faster than light can travel in water, and it sets off a cone-shaped flash of blue light — like a photon boom. Thousands of light sensors on the walls catch that flash, and from its shape physicists can figure out what kind of particle came in and which direction it was going.

Also called: water Cherenkov detector, Cherenkov neutrino detector.

Why it matters

Super-Kamiokande's 1998 detection of a zenith-angle-dependent deficit of atmospheric muon neutrinos provided the first evidence that neutrinos have mass — a result outside the original Standard Model of particle physics.
Water Cherenkov detectors can point back to neutrino sources: Super-Kamiokande was the first experiment to confirm the Sun as a directional neutrino source.
The same technique is sensitive to proton decay, supernova neutrino bursts, and diffuse supernova neutrino background — physics goals spanning the entire lifetime of the universe.
Water is cheap, scalable, and self-purifying with the right systems — allowing kiloton to megaton detectors at a cost per ton far lower than any solid detector.
IceCube has used the ice variant of the technique to identify high-energy astrophysical neutrinos from sources outside the Milky Way, opening a channel of multi-messenger astronomy.

Where it lives

Super-Kamiokande — 50,000-ton ultrapure water tank, 11,129 inward-facing 50 cm PMTs, 1,000 m underground in Mozumi Mine, Japan (operational since 1996)
IceCube Neutrino Observatory — 5,160 digital optical modules on 86 strings, 1 cubic kilometer of Antarctic ice, 1,450–2,450 m depth, South Pole (completed 2010)
SNO (Sudbury Neutrino Observatory) — 1,000 tonnes of heavy water in a 12 m acrylic sphere, ~9,500 PMTs, 2 km underground in Creighton Mine, Canada (1999–2006; solved the solar neutrino problem)
SNO+ — the SNO cavity refilled with liquid scintillator instead of heavy water, located at SNOLAB, Canada (physics data-taking ongoing)
Hyper-Kamiokande — successor to Super-K with 258,000 tons of water and 40,000 PMTs; under construction in Japan, first physics expected ~2027

Key numbers

Super-K water mass: 50,000 tonnes (Ultrapure; inner detector fiducial volume ~22,500 tonnes)
Super-K inner PMTs: 11,129 (50 cm diameter Hamamatsu tubes; ~40% photocathode coverage of inner surface)
IceCube instrumented volume: 1 km³ (86 strings × 60 DOMs; 1,450–2,450 m depth)
Cherenkov angle in water: ~41° (For ultra-relativistic particles; cos θ = 1/n, n ≈ 1.33 for water)
1998 Super-K significance: >5σ (Atmospheric neutrino oscillation result, 33 kton·yr exposure (535 days))

Surprising facts

Super-Kamiokande's 50,000 tonnes of ultrapure water is so clean that light can travel roughly 70 meters through it before being absorbed — far cleaner than any commercial bottled water.
In June 1998, Super-Kamiokande reported that roughly half of all upward-going muon neutrinos — those that had traveled through the entire Earth — were missing compared to predictions, the first experimental evidence that neutrinos change flavor and therefore must have mass.
IceCube's cubic kilometer of instrumented ice sits at depths where the ice is under such pressure that it has been compressed to extraordinary clarity, with a photon absorption length exceeding 100 meters in the clearest layers.
A single solar neutrino detection in Super-Kamiokande requires the faint light from a single electron recoil to be picked up by a detector wall more than 16 meters away — and the experiment catches roughly 15 such events per day.

Open questions

The solar neutrino deficit — 40 years of missing particles: From the late 1960s through 2001, every solar neutrino detector recorded far fewer electron neutrinos from the Sun than solar models predicted — between 30% and 50% of the expected number. Super-Kamiokande measured the solar neutrino flux but could not determine whether the shortfall was a problem with the Sun or with the neutrinos. In 2001–2002, SNO — using heavy water sensitive to all neutrino flavors simultaneously — showed that the total neutrino flux from the Sun matched predictions perfectly; only the flavor composition had changed. The Sun was fine all along. The neutrinos had transformed.
Atmospheric muon neutrinos disappearing with distance through Earth: Super-Kamiokande observed that muon neutrinos arriving from directly above — having traveled only about 15 km from the upper atmosphere — arrived at normal rates. Those arriving from below, having crossed the full ~12,700 km diameter of the Earth, were present at roughly half the expected rate. The Earth itself was acting as a baseline for a neutrino oscillation experiment. The effect was fully consistent with quantum-mechanical flavor oscillation with a mass-squared difference Δm² ~ 2.5 × 10⁻³ eV² — a number that, for more than two decades, the Standard Model had assumed was exactly zero.
IceCube astrophysical neutrinos — bright objects nobody can identify: Since 2013, IceCube has detected a statistically significant flux of very high-energy (TeV–PeV) neutrinos that cannot be explained by atmospheric cosmic-ray backgrounds. Their arrival directions are broadly isotropic, suggesting extragalactic sources. A 2022 analysis found a statistical excess from the direction of the active galaxy NGC 1068 at ~4.2σ significance, and a 2023 result identified a diffuse neutrino emission from the Galactic plane. Despite these tantalizing correlations, the primary sources and acceleration mechanism for the bulk of the astrophysical neutrino flux remain unidentified as of 2026.

A short history

The Cherenkov effect — coherent radiation from charged particles moving faster than light in a medium — was first systematically described by Pavel Cherenkov in 1934 and explained theoretically by Ilya Frank and Igor Tamm; all three shared the 1958 Nobel Prize. Water Cherenkov detectors were pioneered in the 1980s by the IMB and Kamiokande experiments, which detected neutrinos from supernova SN 1987A. Super-Kamiokande, commissioned in 1996 with ten times Kamiokande's volume, announced the atmospheric neutrino oscillation discovery in June 1998 at the Neutrino '98 conference in Takayama, Japan, a result published in Physical Review Letters in August 1998 (Phys. Rev. Lett. 81, 1562). Takaaki Kajita shared the 2015 Nobel Prize in Physics for this discovery. The SNO experiment in Canada subsequently resolved the solar neutrino problem by showing in 2001–2002 that electron neutrinos from the Sun were oscillating into other flavors, with Arthur McDonald sharing the same 2015 Nobel.

Sources

Super-Kamiokande collaboration: Evidence for oscillation of atmospheric neutrinos, Phys. Rev. Lett. 81, 1562 (1998) (https://link.aps.org/doi/10.1103/PhysRevLett.81.1562)
Kajita Nobel Lecture: Discovery of Atmospheric Neutrino Oscillations (2015) (https://www.nobelprize.org/uploads/2018/06/kajita-lecture.pdf)
Super-Kamiokande detector description, Nuclear Instruments and Methods A (ScienceDirect) (https://www.sciencedirect.com/science/article/abs/pii/S016890020300425X)
IceCube detector description — IceCube Neutrino Observatory official site (https://icecube.wisc.edu/science/icecube/)
CERN Courier: SNO — solving the mystery of the missing neutrinos (2007) (https://cerncourier.com/a/sno-solving-the-mystery-of-the-missing-neutrinos/)