Neutrinos are notoriously hard to observe: only 0.001% of those passing through Earth interact with it — making the chance of detecting a neutrino interaction in a particle detector on the order of one per trillion. One would therefore not expect these particles to be the best available probe for obtaining geophysical information regarding our planet’s internal structure. As an analysis by Andrea Donini and colleagues published in this issue of Nature Physics now shows1, however, one would be wrong in jumping to this conclusion.

It turns out that the Earth becomes opaque to neutrinos whose energy exceeds 10 TeV: roughly speaking, the diameter of the Earth represents one absorption length for a neutrino with an energy of 25 TeV. Moreover, neutrinos on such energy scales are produced in collisions of cosmic rays with nuclei in the Earth’s atmosphere. Since their absorption depends on their energy and the distance they have travelled, it is easy to imagine measuring the Earth’s density profile by studying the distribution of TeV atmospheric neutrinos as they pass through our planet.

Indeed, the idea of neutrino tomography is not new and was proposed as long ago as the 1970s2,3. The problem, of course, was the detection of the atmospheric neutrinos themselves. Thankfully, the advent of large-scale neutrino detectors such as the IceCube telescope located in Antarctica has now overcome this issue: significant datasets of the detections of TeV atmospheric neutrinos have been recorded4,5 and, crucially, made publicly available6. As Véronique Van Elewyck describes in her News and Views7, all of this allowed Donini and colleagues1 — who are not directly associated with the IceCube collaboration — to determine the mass of the Earth and its core, its moment of inertia, and indeed establish that the core is more dense than the mantle.

A critic might question what all the fuss is about: after all, this proof of concept is in line with numerical predictions8 and, from a geological perspective, we are hardly learning anything new about the structure of the Earth. But consider this: unlike conventional gravimetric methods currently used in geophysics, the neutrino tomography approach relies solely on the knowledge of weak interactions and of nucleon masses — it is completely distinct, both conceptually and methodologically.

The authors estimate that ten more years of IceCube data will dramatically reduce the statistical errors in their analysis, by which time other neutrino detectors such as KM3NeT (currently under construction in the Mediterranean Sea) will be online, and important geophysical questions could be addressed, such as obtaining an independent verification of the density discontinuities at the core–mantle boundary. Could neutrino tomography turn out to be the geophysical equivalent of magnetic resonance imaging of animal tissue? Physicists and engineers are known to like a challenge, so why not?