Applications
What a Cubic Kilometre of Ice Is For
A detector that records four interesting events a year has to earn its keep by answering questions nothing else can. Each use below is enabled by something derived earlier — and in two cases, limited by it.
1. Multi-messenger astronomy: giving a neutrino an address
In September 2017 a single high-energy neutrino arrived as a track, so by Part 3 its direction was known to about half a degree. An alert went out within a minute, telescopes looked, and in that patch of sky sat a blazar — a galaxy whose supermassive black hole fires a jet nearly along our line of sight — which happened to be flaring in gamma rays at the time.
That is the whole method: one messenger localises, others identify. It only worked because the event was a track rather than a cascade. A cascade would have given a patch of sky fifteen degrees across, containing hundreds of candidate galaxies and settling nothing.
In 2023 the same instrument produced an image of the Milky Way in neutrinos — our own galaxy, seen by a messenger that has never been used to look at it before.
2. Where cosmic rays come from, after a century of not knowing
Charged cosmic rays reach Earth with enormous energies, but they are charged, so galactic and intergalactic magnetic fields bend their paths. By the time one arrives, its direction tells you almost nothing about its origin. The question of what accelerates them has been open since 1912.
Neutrinos solve the pointing problem by construction: neutral, so unbent; weakly interacting, so unabsorbed. Wherever protons are being accelerated they collide with surrounding gas or light and make neutrinos, so a neutrino source is an accelerator caught in the act. This is the argument that made a cubic kilometre worth the money.
3. Particle physics above any accelerator
The ice is also a target for particle physics at energies no machine on Earth will reach. The cleanest example is a resonance whose energy you can derive from two masses.
An electron antineutrino hitting an atomic electron can form a real W boson, if the centre-of-mass energy matches the W mass. For a stationary target electron:
\[ s = 2\,m_e E_{\bar\nu} = m_W^2 \quad\Longrightarrow\quad E_{\bar\nu} = \frac{m_W^2}{2m_e} \]
With mW = 80.4 GeV and me = 0.511 MeV:
\[ E_{\bar\nu} = \frac{(80.4)^2}{2 \times 5.11\times10^{-4}}\ \text{GeV} \approx 6.3\ \text{PeV} \]
Predicted by Sheldon Glashow in 1960, and for sixty years a curiosity, because nothing could produce an electron antineutrino at 6.3 PeV. In 2021 IceCube reported a candidate event at that energy — a prediction from the earliest days of electroweak theory, tested with a particle from another galaxy and a block of Antarctic ice.
The same instrument has measured the neutrino–nucleon cross-section above a TeV by a neat trick: counting how many neutrinos the Earth absorbs as a function of arrival angle. The planet becomes the apparatus.
4. Waiting for a supernova
A core-collapse supernova in our galaxy would release almost all its energy as neutrinos of a few tens of MeV — far too low for any of the reconstruction in Part 3 to work. IceCube would not see tracks or cascades at all.
What it would see is every one of its 5,160 sensors brighten together for a few seconds, as the whole ice sheet glows faintly. It cannot say where the supernova is, only that one is happening — which, broadcast within seconds, is enough for every other telescope to start looking. A kilometre-scale detector built for PeV neutrinos turns out to double as an alarm clock for the MeV ones.
5. Glaciology, by accident
To reconstruct events at all, the collaboration had to measure how light scatters and is absorbed through a kilometre of ice, layer by layer. The result was an unintended high-resolution record of the Antarctic ice sheet: dust bands from ancient volcanic eruptions and climate shifts, laid down over a hundred thousand years and still legible. The instrument had to understand the glacier to do astronomy, and ended up contributing to glaciology instead.
What is still hard
Almost everything is statistics-limited: a handful of astrophysical events per year, from Part 2's arithmetic, means most individual sources remain unidentified and the diffuse flux dominates. Only muon tracks point well, and they are a minority of events. Above a few PeV the Earth itself becomes opaque, closing the downward sky exactly where the physics is most interesting. The next step is simply more volume — which is why planned detectors are measured in tens of cubic kilometres.