Neutrino Atlas
EN
  1. Learn
  2. Neutrino astronomy

Chapter 4: Neutrinos in the universe · Lesson 4.2

Neutrino astronomy

Confirmed

Neutrinos cross the Universe in straight lines, unabsorbed and undeflected. Since 2013 giant detectors have used them to locate cosmic particle accelerators, work recognised by the 2026 Nobel Prize in Physics.

Event shapes in a neutrino telescopeThree patterns of lit sensors: a long straight track, a round cascade, and two cascades connected by a short track (double bang).Trackν_μ → muon; good directionCascadeν_e, NC; good energyDouble bangν_τ → τ; at PeV energies
In IceCube, KM3NeT and Baikal-GVD the flavour of a high-energy neutrino shows in the pattern of light. A muon from a muon-neutrino interaction travels kilometres and leaves a long track that points back to the source within about a degree. Electron neutrinos and all neutral-current interactions produce compact, almost spherical cascades whose energy is measured well but whose direction is less precise. A tau neutrino can produce two cascades separated by the flight of the tau lepton – a signature IceCube identified in 2024. Schematic, sensor brightness indicated by size.

Beginner

A new kind of telescope

For most of history, astronomy meant collecting light: first visible light, later radio waves, X-rays and gamma rays. Neutrino astronomy collects something else. A neutrino is a tiny, electrically neutral particle that hardly ever interacts with anything. Billions of them pass through your thumbnail every second without leaving a trace. That makes them very hard to catch, but it also makes them unusually honest messengers from space.

Why neutrinos make good messengers

Light from a distant galaxy can be blocked by dust, absorbed by gas or scattered on its way to us. Charged particles from space, called cosmic rays, are bent by magnetic fields, so by the time they arrive their direction no longer points back to where they came from. Neutrinos have neither problem. They have no electric charge, so magnetic fields do not bend them, and they pass through dust, gas and even whole stars. A neutrino that arrives at Earth travels in a straight line from its source, like a letter that is never opened or redirected.

Very energetic neutrinos are made where nature accelerates particles to extreme speeds, for example near giant black holes. When fast protons collide with gas or light, they produce short-lived particles that decay into neutrinos. Finding such neutrinos tells us where these natural particle accelerators are.

Building a detector out of ice and water

Because neutrinos so rarely interact, a detector must be enormous. Physicists found a clever solution: use a natural block of clear water or ice and fill it with light sensors. When a neutrino does hit an atom, it produces a fast charged particle that emits a faint blue glow called Cherenkov light. The sensors record this glow, and its pattern reveals the direction and energy of the neutrino.

Early attempts included DUMAND in the deep Pacific off Hawaii, which was cancelled in 1995, a detector in Lake Baikal in Siberia, and AMANDA in the Antarctic ice. Their successor, IceCube at the South Pole, uses one cubic kilometre of ice, roughly the volume of a small mountain, watched by more than five thousand sensors.

What has been found

  • 2013: IceCube found the first convincing evidence that very energetic neutrinos reach us from outside our Solar System1.
  • 2017–2018: a single neutrino was traced back to a distant galaxy called TXS 0506+056, which was flaring at the same time2.
  • 2022: neutrinos were seen coming from the nearby galaxy NGC 10683.
  • 2023: the band of our own Milky Way was seen for the first time in neutrinos rather than light4.
  • 2025: the European KM3NeT detector in the Mediterranean Sea reported the most energetic neutrino ever recorded6.

Recognition and the next step

In October 2026, Francis Halzen, the Belgian-born physicist who led the AMANDA and IceCube projects from the University of Wisconsin–Madison, received the Nobel Prize in Physics as the sole laureate8. The prize marks neutrino astronomy as a working branch of astronomy rather than a hope.

The field is still young. Only a handful of sources are known, and most cosmic neutrinos cannot yet be traced to a home. Larger detectors are being built in Lake Baikal and the Mediterranean, and a much larger version of IceCube, called IceCube-Gen2, has been proposed. Together they should turn today's first sketches of the neutrino sky into a detailed map.

Quiz

Test your knowledge

4 questions on this level · Print as worksheet

  1. 1Why do neutrinos point back to their source while cosmic rays do not?
  2. 2What do the sensors in IceCube record?
  3. 3How large is the IceCube detector?
  4. 4What was seen for the first time in 2023?

Intermediate

Messengers that arrive unchanged

High-energy astrophysical neutrinos are produced when accelerated protons or nuclei collide with matter or radiation near their source. These collisions create charged pions, which decay into muons and neutrinos; neutral pions decay into gamma rays at the same time. Neutrinos are therefore the cleanest tracers of hadronic cosmic-ray acceleration. Unlike gamma rays, they are not absorbed by dense gas or by the extragalactic background light, and unlike cosmic rays they are not deflected by magnetic fields. Their weakness is their tiny cross-section: even at TeV energies only a minute fraction interacts on the way through a detector.

From an idea to a cubic kilometre

In 1960 Moisey Markov proposed observing neutrino-induced muons with detectors deep in a lake or the sea. The DUMAND project off Hawaii, developed from 1976 by John Learned, Arthur Roberts and others, worked out the technology but was cancelled in 1995. The Baikal collaboration deployed NT-36 in Lake Baikal in 1993 and recorded the first underwater neutrino-induced muons. At the South Pole, AMANDA placed strings of photomultipliers in the ice from 1993/94 and showed that deep Antarctic ice is clear enough for the method.

IceCube was built between 2005 and 2010. It consists of 5,160 digital optical modules on 86 strings between 1,450 and 2,450 m below the surface, instrumenting about one cubic kilometre. A muon neutrino produces a long muon track that points back to the source within about a degree; electron and most tau neutrinos produce compact showers (cascades) whose energy is measured well but whose direction is poorly known.

cos θC = 1 / (n β)Cherenkov angle θC for a particle of speed β (in units of the speed of light) in a medium of refractive index n. In ice and water (n ≈ 1.33) relativistic particles emit at about 41°, which allows the track direction to be reconstructed from photon arrival times.

Milestones

  • 2013: 28 events between about 30 TeV and over 1 PeV, starting inside IceCube, were inconsistent with atmospheric backgrounds at 4.1σ. This was the first evidence for an astrophysical neutrino flux1.
  • 2017–2018: the alert IceCube-170922A, a muon neutrino of about 290 TeV, coincided with a gamma-ray flare of the blazar TXS 0506+056 at about 3σ; an archival search found an independent 3.5σ excess from the same direction in 2014–20152.
  • 2021: a shower of about 6.05 PeV visible energy matched the Glashow resonance, predicted by Sheldon Glashow in 19605.
  • 2022: an excess of about 79 neutrinos from the active galaxy NGC 1068 at 4.2σ, the first evidence for steady emission from a single source3.
  • 2023: neutrino emission from the plane of the Milky Way at 4.5σ, found with ten years of cascade events4.
  • 2025: KM3NeT reported KM3-230213A, a muon whose energy implies a neutrino of about 220 PeV6.

A global network

In the Mediterranean, KM3NeT is building two arrays: ARCA off Sicily for high-energy astronomy and ORCA off Toulon for oscillation physics. In Lake Baikal, Baikal-GVD is growing cluster by cluster and has reported its own evidence for the diffuse astrophysical flux7. Because neutrino telescopes see best through the Earth, detectors in both hemispheres cover complementary parts of the sky. IceCube sends real-time alerts to optical, X-ray and gamma-ray observatories, a practice central to multimessenger astronomy. The proposed IceCube-Gen2 would enlarge the optical array to about eight cubic kilometres.

The 2026 Nobel Prize

On 6 October 2026 the Royal Swedish Academy of Sciences awarded the Nobel Prize in Physics to Francis Halzen of the University of Wisconsin–Madison, as sole laureate, for decisive contributions to IceCube and the discovery of high-energy neutrinos of astrophysical origin8. Halzen had promoted Antarctic ice as a detection medium and served as principal investigator of AMANDA and IceCube.

Quiz

Test your knowledge

4 questions on this level · Print as worksheet

  1. 1Why are neutrinos the cleanest tracers of hadronic cosmic-ray acceleration?
  2. 2Why do muon tracks give better directions than cascades?
  3. 3At about what angle do relativistic particles emit Cherenkov light in ice or water?
  4. 4What did KM3NeT report in 2025?

Expert

Production and the multimessenger connection

In hadronic sources, accelerated protons interact with ambient gas (pp) or photon fields (pγ). Both channels produce pions; the decay chain π⁺ → μ⁺ νμ, μ⁺ → e⁺ νe ν̄μ gives a source flavour ratio νe : νμ : ντ ≈ 1 : 2 : 0, which neutrino mixing over cosmic baselines averages to close to 1 : 1 : 1 at Earth (astrophysical flavour ratio). Each neutrino carries roughly 5% of the parent proton energy, so PeV neutrinos point to accelerators of tens of PeV. The accompanying π⁰ → γγ photons link the neutrino and gamma-ray luminosities; the key diagnostic of recent years is that several neutrino sources are underluminous in GeV–TeV gamma rays, implying sources opaque to their own photons.

dN/dE ∝ E^(−γ)Power-law spectrum of the diffuse flux. Fits to IceCube data give indices γ of roughly 2.4–2.9 depending on event sample and energy range; Baikal-GVD finds γ ≈ 2.6 for its cascade sample. The per-flavour normalisation near 100 TeV is of order 10⁻¹⁸ GeV⁻¹ cm⁻² s⁻¹ sr⁻¹.

Detection and event classes

IceCube's 5,160 optical modules on 86 strings (125 m spacing, with the denser DeepCore infill) observe Cherenkov photons in ice whose effective scattering length is a few tens of metres. Throughgoing muon tracks reach median angular resolutions below 1° at high energy but only lower bounds on neutrino energy; contained cascades give ~10–15% energy resolution with directions of order 10°. Starting tracks combine both, and the outer-layer self-veto suppresses atmospheric neutrinos from the southern sky that arrive with their parent air shower. A double-bang topology is the signature of a ντ; IceCube has reported a small number of tau candidates.

Results to date

  • Diffuse flux: 28 high-energy starting events in two years, 4.1σ above atmospheric backgrounds1; the signal has since been confirmed in throughgoing muons, cascades and by Baikal-GVD at 5.1σ with upgoing cascades from 2018–20247.
  • TXS 0506+056: IceCube-170922A (~290 TeV) coincident with a Fermi-LAT and MAGIC flare at ≈3σ, plus a 2014–2015 archival excess of 13 ± 5 events at 3.5σ2.
  • Glashow resonance: a 6.05 ± 0.72 PeV visible-energy cascade recorded in December 2016, consistent with ν̄e e⁻ → W⁻ at 2.3σ, showing that the flux contains antineutrinos5.
  • NGC 1068: 79 (+22/−20) events at 4.2σ from a Seyfert galaxy faint in TeV gamma rays, pointing to emission from the X-ray-bright corona close to the black hole3.
  • Galactic plane: 4.5σ evidence from ten years of cascades analysed with machine-learning selection; the flux is comparable to the expectation from cosmic rays interacting with interstellar gas4.
  • KM3-230213A: a nearly horizontal muon of about 120 PeV in the partially built ARCA detector, implying Eν ≈ 220 PeV (central value; 90% range roughly 72 PeV to 2.6 EeV); the absence of comparable IceCube and Auger events creates a mild tension6.
Eres = mW² / (2 me) ≈ 6.3 PeVNeutrino energy for resonant W⁻ production on an atomic electron at rest (Glashow resonance), with mW the W-boson mass and me the electron mass. Only ν̄e participate, so the rate tests the neutrino–antineutrino ratio of the source population.

Source populations and the energy frontier

The diffuse flux is far larger than the identified sources can explain. Stacking analyses limit gamma-ray-bright blazars to a minority contribution, and the isotropic gamma-ray background measured by Fermi-LAT constrains transparent pp sources, because their π⁰ photons would cascade down into the GeV band. Both arguments favour populations that are dark in gamma rays, such as cores of active galaxies like NGC 1068. Above about 100 PeV the expected flux is dominated by cosmogenic neutrinos from ultra-high-energy cosmic rays interacting with the cosmic microwave background, a prediction going back to Berezinsky and Zatsepin in 1969. Reaching it requires instrumented volumes of hundreds of cubic kilometres, which is why radio detection in ice is being developed alongside optical arrays.

Detectors and the Nobel Prize

The lineage runs from Markov's 1960 proposal through DUMAND (cancelled 1995), Baikal NT-36/NT-200 (1993/1998), AMANDA (1993–2009) and ANTARES (2008–2022) to IceCube (completed December 2010). KM3NeT uses multi-PMT digital modules in two arrays: ARCA (3,450 m deep, off Sicily) and ORCA (off Toulon). Baikal-GVD had 14 clusters with 4,212 optical modules in 2025, an effective cascade volume of about 0.7 km³ at PeV energies. The IceCube Upgrade added five densely instrumented strings in the 2025/26 season for calibration and low-energy physics, and IceCube-Gen2 would add an optical array of about 8 km³ and a radio array for EeV energies.

The 2026 Nobel Prize in Physics went to Francis Halzen alone, for decisive contributions to the IceCube Neutrino Observatory and the discovery of high-energy neutrinos of astrophysical origin8. It was the first undivided physics prize since Georges Charpak's in 1992.

Quiz

Test your knowledge

4 questions on this level · Print as worksheet

  1. 1What flavour ratio νe : νμ : ντ does pion decay give at the source, and at Earth?
  2. 2Which neutrinos take part in the Glashow resonance at Eres = mW²/(2me) ≈ 6.3 PeV?
  3. 3PeV neutrinos point to accelerators of roughly what proton energy?
  4. 4Why does the Fermi-LAT isotropic gamma-ray background favour gamma-dark neutrino sources?

Open questions

  • Which source classes produce the bulk of the diffuse flux? Identified sources such as TXS 0506+056 and NGC 1068 account for only a small fraction.
  • Is KM3-230213A cosmogenic, from an extreme source, or a statistical fluctuation given IceCube's non-detection at similar energies?
  • Do the measured flavour composition and neutrino–antineutrino ratio agree with standard pion production and three-flavour mixing?

Terms to explore

People

On the timeline

  1. 1960Markov proposes deep underwater neutrino detection
  2. 1976DUMAND plans a deep-ocean neutrino telescope
  3. 1993First underwater neutrino telescope in Lake Baikal
  4. 1994AMANDA deploys detectors in the South Pole ice
  5. 2010IceCube is completed at the South Pole
  6. 2013IceCube finds high-energy astrophysical neutrinos
  7. 2018Blazar TXS 0506+056 linked to a neutrino
  8. 2021IceCube detects a Glashow resonance candidate
  9. 2022IceCube sees neutrinos from galaxy NGC 1068
  10. 2023IceCube observes neutrinos from the Milky Way
  11. 2025KM3NeT detects a 220 PeV neutrino
  12. 2026IceCube Upgrade strings deployed
  13. 2026Nobel Prize for Francis Halzen and IceCube

References

  1. IceCube Collaboration, M. G. Aartsen et al., 'Evidence for high-energy extraterrestrial neutrinos at the IceCube detector', Science 342, 1242856 (2013), doi:10.1126/science.1242856
  2. IceCube, Fermi-LAT, MAGIC et al., 'Multimessenger observations of a flaring blazar coincident with high-energy neutrino IceCube-170922A', Science 361, eaat1378 (2018); IceCube Collaboration, 'Neutrino emission from the direction of the blazar TXS 0506+056 prior to the IceCube-170922A alert', Science 361, 147 (2018), doi:10.1126/science.aat1378, doi:10.1126/science.aat2890
  3. IceCube Collaboration, R. Abbasi et al., 'Evidence for neutrino emission from the nearby active galaxy NGC 1068', Science 378, 538 (2022), doi:10.1126/science.abg3395
  4. IceCube Collaboration, R. Abbasi et al., 'Observation of high-energy neutrinos from the Galactic plane', Science 380, 1338 (2023), doi:10.1126/science.adc9818
  5. IceCube Collaboration, M. G. Aartsen et al., 'Detection of a particle shower at the Glashow resonance with IceCube', Nature 591, 220 (2021), doi:10.1038/s41586-021-03256-1
  6. KM3NeT Collaboration, S. Aiello et al., 'Observation of an ultra-high-energy cosmic neutrino with KM3NeT', Nature 638, 376 (2025), doi:10.1038/s41586-024-08543-1
  7. Baikal-GVD Collaboration, 'Measurement of the diffuse astrophysical neutrino flux over six seasons using cascade events from the Baikal-GVD expanding telescope', arXiv:2507.01893 (2025), arXiv:2507.01893
  8. The Royal Swedish Academy of Sciences, 'The Nobel Prize in Physics 2026', press release, nobelprize.org (6 October 2026), nobelprize.org/prizes/physics/2026/summary/

Last updated: · Editorial standard

Cite this page

Text

Neutrino Atlas (2026). “Neutrino astronomy”. Retrieved 22 November 2026, from https://neutrino-atlas.com/learn/neutrino-astronomy/

BibTeX

@misc{neutrinoatlas2026learnneutrinoastronomy,
  title        = {Neutrino astronomy},
  author       = {{Neutrino Atlas}},
  year         = {2026},
  url          = {https://neutrino-atlas.com/learn/neutrino-astronomy/},
  note         = {Accessed 2026-11-22},
  language     = {english}
}