
Can a Mobile Neutrino Detector in the Deep Pacific Map Earth’s Hidden Heat?
An arXiv white paper proposes a relocatable ocean-bottom antineutrino detector that, based on models, could reduce crustal background 50–100-fold relative to similarly sized continental detectors and enable direct measurements of mantle geoneutrinos, probing Earth’s radiogenic heat budget.
Jun 18, 2026
Somewhere in the middle of the Pacific Ocean, more than four kilometers below the surface, a detector could sit in the dark and catch particles that have passed through the entire Earth. Those particles are geoneutrinos—antineutrinos produced in radioactive decay chains, especially uranium and thorium. They rarely interact, which is why they can escape the deep interior and why they are so hard to catch. A new white paper argues that putting a detector in the deep ocean, far from continents, may be the best way to hear them.
The arXiv white paper is a concept and prototyping study, not a detection claim. It describes the Ocean Bottom Detector (OBD): a mobile, kiloton-scale liquid-scintillator experiment designed to measure geoneutrinos from Earth’s mantle. The goal is to answer a question that surface rocks and seismic waves cannot settle directly: How much of Earth’s internal heat comes from radioactive decay, and where are those heat-producing elements stored?
An Earthly Heat Budget With a Missing Term
Earth loses about 46 ± 3 terawatts (TW, trillion watts) of heat through its surface. The continental crust contributes about 7 ± 1 TW of radiogenic heating, from decay of uranium, thorium, and potassium. The remaining ~39 TW is the focus of debate, especially how much comes from radiogenic heating in the mantle versus primordial heat left over from Earth’s formation.
The mantle term is the problem. Published estimates for mantle radiogenic power range from about 2 to ≥20 TW. Different compositional models for the bulk Earth predict total radiogenic heating anywhere from about 10 to 33 TW. That uncertainty matters because radiogenic heat helps power mantle convection, plate tectonics, volcanism, and the geodynamo that shields the planet.
Geoneutrinos offer a direct check. The uranium-238 and thorium-232 decay chains emit antineutrinos through beta decays, and those particles travel through the planet almost unimpeded. Detecting them and measuring their energies can, in principle, reveal how much uranium and thorium are decaying in the Earth today.
Why the Deep Pacific Changes the View
Since KamLAND’s first geoneutrino observation in 2005, Borexino, SNO+, and JUNO have also reported geoneutrino measurements. JUNO’s contribution was extracted in its first reactor-analysis fit; the paper notes that a dedicated geoneutrino flux analysis has not yet been reported. All of these detectors are underground liquid-scintillator experiments on continental crust. They catch antineutrinos through inverse beta decay: an antineutrino strikes a proton, producing a positron and a neutron. The positron’s light and the later neutron capture form a coincidence signal.
Continental detectors are swamped by nearby crust. At KamLAND, roughly 70 percent of the geoneutrino flux comes from the crust, with only about 30 percent from the mantle. The local crust’s composition is uncertain, and nuclear reactors produce antineutrinos that complicate the measurement. The mantle signal is there, but it is not cleanly isolated.
The white paper proposes moving the detector offshore. In the deep ocean, far from continents, the oceanic crust is only about 6 to 7 kilometers thick, compared with 30 to 40 kilometers for continental crust. It is also depleted in heat-producing elements. The authors estimate that this setting could reduce the crustal geoneutrino background by 50 to 100 times compared with a similarly sized continental detector.
At a mid-Pacific site such as ALOHA, north of Oahu, modeled spectra predict that about 70 percent of the geoneutrino signal would come from the mantle. The detector would also sit beneath 4 to 5 kilometers of water-equivalent overburden, which shields it from cosmic rays, and far from most nuclear reactors.
The OBD design uses a liquid scintillator inside a stainless-steel tank and an acrylic vessel. A buffer oil separates the scintillator from inward-facing photosensors. The paper’s table gives target masses of 1–20 kilotons, and specific simulations compare 1.5- and 17-kiloton configurations, with deployment depths around 4 kilometers.
Mobility is the key difference. A relocatable detector could be parked at different ocean sites, gathering multiple stereoscopic views of the geoneutrino flux. That could turn a single mantle average into a map of uranium and thorium variations. With directional sensitivity—enhanced by doping the scintillator with lithium or gadolinium and using pattern-matching algorithms—the detector might even probe large low-velocity provinces, 1,000-kilometer-tall structures at the core-mantle boundary whose origin and composition remain debated.
Prototype work is already underway. At Tohoku University, researchers built a 12-liter prototype for optical tests and ran a separate low-temperature study of a liquid scintillator. Near a concentration of 3 grams per liter of the fluor PPO, the scintillator’s light yield at 4 degrees Celsius was about 8.6 percent higher than at 20 degrees Celsius, after correcting for the photomultiplier’s own temperature response. Structural studies examined a 1.5-kiloton geometry under deep-sea pressure and creep in acrylic; simulations also compared 1.5- and 17-kiloton configurations. At the University of Hawai‘i, researchers are testing photosensors and are interested in using the ALOHA Cabled Observatory, 4.7 kilometers deep, for in-situ tests.
Beyond Earth’s Interior
The same detector could serve other physics. If a core-collapse supernova exploded within about 10 kiloparsecs, an OBD-like instrument could record several hundred neutrinos per second; SN 1987A yielded only about 19 detected neutrinos in total. That would let astronomers watch the accretion plateau, the cooling phase, and perhaps an abrupt cutoff signaling black hole formation.
The paper also notes potential uses for nuclear monitoring. A mobile ocean-bottom detector could monitor coastal reactors or future floating nuclear plants from just outside territorial waters. But the authors caution that a reactor-optimized detector would complicate precision geoneutrino measurements because reactor antineutrinos dominate, and that near continental margins the local crustal geoneutrino flux dominates; legal and diplomatic limits would also shape such missions.
The main caveat is that mantle geoneutrinos have not yet been measured directly. The projected 50-to-100-fold background reduction comes from simulations and geoscientific models, not from a deployed full-scale OBD. Deep-ocean operations, pressure tolerance, radiopurity, cost, and data analysis still need to be demonstrated. The white paper is a roadmap, not a result.
Still, the idea reverses the usual problem in neutrino astronomy. Instead of looking outward, it would use some of nature’s most elusive particles to look inward—turning the ocean into a telescope aimed at the Earth’s hidden heat.