The Jiangmen Underground Neutrino Observatory (JUNO) published its first physics data on June 10, 2026, in Nature — as a cover story. The result is a measurement of neutrino oscillations with a precision 1.6 times greater than all previous experiments combined. This is no marginal update: it is the largest qualitative leap in neutrino physics in decades.
The data collected span just 59 days, between August 26 and November 2, 2025. During that period, the detector captured antineutrinos emitted by two nuclear power plants in China’s Guangdong province, roughly 53 kilometers away. JUNO is buried 700 meters underground and contains 20,000 tonnes of liquid scintillator — the largest sensitive mass in the world for this type of instrument, equipped with over 17,000 twenty-inch photomultiplier tubes. The nearby nuclear reactors are no coincidence: they produce abundant, steady, and well-characterized antineutrino fluxes, making them near-ideal sources for oscillation physics.
Seventy years ago, physicists Clyde Cowan and Frederick Reines built a 10-tonne detector, surrounded it with lead walls and sandbags soaked in water, and placed it next to a nuclear reactor at Savannah River, South Carolina. They called the experiment Project Poltergeist. In 1956, they achieved the first direct detection of the neutrino. That logic — using a reactor as a source, shielding the detector from background noise, and waiting for an extremely rare signal — has remained unchanged. What has changed is the scale.
JUNO has confirmed the so-called “solar neutrino tension,” a discrepancy of roughly 1.5 sigma between oscillation parameter measurements that has persisted for years across different experiments. It does not yet formally contradict the standard model of particle physics, but it signals that something may not add up. The observatory’s primary goal remains to determine the neutrino mass hierarchy — that is, to establish which of the three neutrino “families” is the heaviest — an open problem for decades with direct implications for our understanding of the matter-antimatter asymmetry in the universe.
JUNO does not work alone. It is part of a global network that includes IceCube at the South Pole, Super-Kamiokande in Japan, KM3NeT on the Mediterranean seabed, and the future DUNE experiment. IceCube uses Antarctic ice as its detection medium to search for very high-energy neutrinos produced by cosmic sources. KM3NeT anchors nearly 200,000 optical sensors to the ocean floor, at a depth of 3,500 meters. Each instrument covers a different energy window; together, they build a picture of neutrino physics that no single experiment could achieve on its own.
Nuclear reactors have made all of this physics possible. From Cowan and Reines’ makeshift trap to JUNO, they have remained the most powerful and controllable artificial neutrino source available. Every new milestone in our understanding of these particles has been built, in large part, on their capacity to produce measurable antineutrino fluxes. With JUNO’s data continuing to accumulate, and experiments like DUNE still under construction, the coming years may finally answer the questions that particle physics has carried open since Reines telegraphed Pauli in 1956: “We have found the ghost.”
