Super-Kamiokande reports 2.6σ hint of diffuse supernova neutrino background

The Super-Kamiokande Collaboration has reported a 2.6-sigma indication of the diffuse supernova neutrino background in a newly posted preprint, offering the strongest experimental hint yet of a signal physicists have sought for decades. But the result, posted on arXiv on Oct. 27, 2026, stops well short of a discovery: in particle and neutrino physics, researchers typically require 5 sigma to claim a definitive detection.

The signal in question is the diffuse supernova neutrino background, or DSNB — the faint, all-sky flux of neutrinos and antineutrinos produced by all past core-collapse supernovae across cosmic history. If firmly detected, it would give astronomers a new way to study the long record of exploding stars and stellar collapse in the universe, not by looking at light from individual events but by measuring the accumulated neutrino glow they leave behind.

In the preprint, titled “Indication of Electron Antineutrinos from the Diffuse Supernova Neutrino Background,” the collaboration reports evidence for electron antineutrinos based on a 5003.0-day dataset from the Super-Kamiokande detector in Japan. The data span October 2008 to March 2026 and combine both pure-water running and later gadolinium-loaded phases of the detector. Using a representative DSNB model, the team found a best-fit antineutrino flux of 3.6^{+1.6}_{-1.5} per square centimeter per second for neutrino energies above 13.3 MeV. That fit corresponds to a 2.6-sigma tension with the no-flux hypothesis, or 99.5% confidence level on a one-sided test. The search covered a reconstructed energy range corresponding to neutrino energies from 13.3 to 81.3 MeV.

Super-Kamiokande searches for these antineutrinos mainly through inverse beta decay in water: an antineutrino interacts with a proton, producing a positron and a neutron. The neutron is the crucial second signature, and it is easier to identify when the detector water contains gadolinium, an element that greatly improves neutron-capture detection. The collaboration said its improved sensitivity also came from stronger rejection of atmospheric neutrino backgrounds, better treatment of spallation and lithium-9 backgrounds, updated estimates of systematic uncertainties, and a refined spectral fitting method. It also said similar significance levels were obtained with different DSNB spectral models, suggesting the result does not depend narrowly on one assumed signal shape.

Hiroyuki Sekiya, Super-Kamiokande spokesperson, called the result an important step while stressing its limits. “Observing the world’s first indication of the Diffuse Supernova Neutrino Background is a deeply meaningful achievement ... At the same time, the current statistical significance is 2.6 sigma, and additional data accumulation and further improvements in analysis are essential for definitive detection. We will continue observations at Super-Kamiokande and work toward a conclusive detection,” he said in a June 26, 2026, press release from the University of Tokyo’s Institute for Cosmic Ray Research. That caution is central to the claim: 2.6 sigma is a meaningful hint, but not the level the field treats as settled.

The collaboration had previewed the result at the Neutrino 2026 conference on June 25, 2026, and the institute announced it in a press release the next day, before the full paper appeared on arXiv. The new analysis marks a shift from earlier Super-Kamiokande searches, including a 2021 study that set upper limits rather than reporting a signal. For now, the new paper stands as the strongest sign yet that the background from all past stellar explosions may finally be coming into view — but confirmation will require more data and continued scrutiny.