Geoneutrinos

The Earth emits heat at roughly 47 TW at its surface. Approximately half of this heat is attributed to radioactive decay of long-lived isotopes in the crust and mantle — principally ${}^{238}$U, ${}^{232}$Th, and ${}^{40}$K. Each decay chain produces electron antineutrinos alongside the heat; measuring the antineutrino flux therefore provides a direct, model-independent probe of terrestrial radiogenic heat.

Sources and spectrum

The dominant ${\bar{\nu }}_e$ emitters are the beta-decaying nuclides in the ${}^{238}$U and ${}^{232}$Th chains. Both chains include multiple ${\beta }^{-}$ transitions with individual Q-values up to ~3.3 MeV. Only antineutrinos from ${}^{214}$Bi and ${}^{234}$Pa in the U chain, and ${}^{228}$Ac and ${}^{212}$Bi in the Th chain, exceed the inverse-beta-decay threshold of 1.806 MeV and are therefore detectable.

${}^{40}$K decays through two channels: ${\beta }^{-}$ to ${}^{40}$Ca (below IBD threshold, invisible) and electron capture to ${}^{40}$Ar (no IBD signal). Geoneutrinos from ${}^{40}$K are therefore invisible to scintillator detectors; their existence contributes to the heat budget but not to the measured flux.

The total predicted ${\bar{\nu }}_e$ flux at Earth’s surface is approximately $6\times 10^6$ cm⁻²s⁻¹ above the IBD threshold, dominated by the nearest 500 km of crust — the “local geology” contribution — with the remainder from the whole Earth.

Detection

Geoneutrinos are observed by large, low-background, liquid-scintillator detectors through IBD. Two experiments have definitively detected them:

KamLAND, at the Kamioka site on continental crust, reported the first positive detection in 2005 and has since accumulated a significant event sample. The measured rate is consistent with predictions from bulk-silicate-Earth models, with $\sim$20% of the total flux attributed to the local continental crust.

Borexino, at Gran Sasso also on continental crust, has published measurements consistent with KamLAND and confirms the mantle contribution to Earth’s heat budget.

SNO+ and JUNO, with much larger target masses, will reduce statistical uncertainties by an order of magnitude.

A detector on oceanic crust — which would be thinner and hence give a cleaner mantle signal — remains a long-standing proposal; the HANOHANO concept would place such a detector on the Pacific seafloor.

What the data tell us about Earth

Combining KamLAND and Borexino and comparing with bulk-Earth models, several conclusions emerge:

1. Half of Earth’s heat flow is radiogenic, within factor-two uncertainty. The remainder is primordial heat from planetary formation and core solidification.
2. The bulk-silicate-Earth model is consistent. Measured rates match models that distribute U and Th through the mantle at abundances inferred from chondritic meteorites.
3. The Th/U ratio in the mantle appears consistent with the canonical value of ~3.9, supporting a primitive-mantle origin for this component.
4. There is no reactor operating in the core. The hypothesis of a natural uranium reactor at the core-mantle boundary, occasionally proposed, is disfavored by the data.

Future reach

Larger detectors and detectors at different sites will enable tomographic geoneutrino maps. Differences in measured rate between continental and oceanic sites separate crustal from mantle contributions. High-statistics spectral measurements may in principle distinguish U from Th decay contributions.

Geoneutrinos sit at an interesting intersection of particle physics, geoscience, and planetary science. They are the only direct electromagnetic-independent probe of the composition of the Earth’s interior.