On October 21, a research team including the National Institute of Advanced Industrial Science and Technology (AIST) announced that it had confirmed numerous normal faults spreading across parts of the southern Okinawa Trough where relatively little heat flows out of the ground. In the formation of ocean basins, the heat from magma has been seen as important for weakening the crust. These observations suggest that the crust can be stretched even where evidence of large-scale magma supply is scarce. However, low heat flow does not prove that magma is absent. A survey combining heat and subsurface structure offers a view of the seafloor at the stage when it begins to split.
Why Study Heat and Faults Together
The Okinawa Trough is a back-arc basin, where the crust is being stretched behind the Ryukyu Arc. A back-arc basin is an ocean basin formed when the upper crust is pulled apart near a site where a plate is subducting. The targets of this study are the Yaeyama Graben and the Yonaguni Graben, which are in the early stage before the transition to full-scale seafloor spreading.
It is thought that when magma rises from below, its heat lowers the strength of the crust and makes it easier to deform. On the other hand, normal faults, in which rock pulled apart breaks and one side slips downward, also accommodate crustal extension. How heat and faults each contribute when the formation of a rift, an elongated depression, begins remains a matter of debate.
The paper was compiled by Makoto Otsubo of AIST, Masataka Kinoshita of the University of Tokyo, Minoru Ikehara of Kochi University, Ryuta Arai of the Japan Agency for Marine-Earth Science and Technology, and others. It was published on October 21, 2026, in Progress in Earth and Planetary Science as a peer-reviewed original paper (DOI: 10.1186/s40645-026-00854-3). Its distinguishing feature is that, rather than discussing faults, heat, and magnetism separately, it compares the two grabens within a common observational framework.
Low heat flow itself had been reported earlier in the Yaeyama Graben. What is new here is that the study links that thermal environment to how faults are distributed in the shallow subsurface. The involvement of magma, which is hard to judge from heat alone, can also be examined by overlaying information on subsurface structure and magnetism.
Heat and Faults Measured in Two Grabens
The study used data from two cruises of the research vessel Hakuho Maru and one cruise of the Shinsei Maru. The main seismic reflection survey was carried out during a Hakuho Maru cruise from December 2023 to January 2024. The paper analyzes in detail two representative cross-sections: line YA4, about 10 km long on the Yaeyama side, and line YO2, about 30 km long on the Yonaguni side.
Seismic reflection surveying sends artificial vibrations from near the sea surface and records the waves that bounce back from boundaries between layers and other features. Instead of drilling underground to see faults, it reads structure from the wave records. In the resulting cross-sections, numerous normal faults cutting the seafloor were recognized in both grabens, and on the Yonaguni line they were distributed over at least about 20 km. In Yaeyama, more densely spaced faults form a stair-step structure.
For the heat measurements, seven temperature sensors were attached to a 4-meter sediment sampler to measure how temperature changes with depth. The researchers also measured how easily heat travels through the collected sediment, and multiplied the two to obtain crustal heat flow. The paper lists five sites as newly acquired measurements, and the uncertainty in heat flow is given as roughly 10%. This is the measurement uncertainty, not a rate of change from earlier values.
Heat flow at most sites was 40 to 63 mW/m² (40 to 63 milliwatts per square meter). However, a locally high value of about 730 mW/m² was obtained at HPC26, near the axis of a graben. The team suggested that this exception may reflect hydrothermal circulation or the influence of local magma. Widespread low values and the existence of sites with high values are compatible.
These are observations obtained at sea and geological interpretations based on those records. They are not the results of a simulation that stretched the crust, nor an experiment that manipulated temperature to test a cause.
What Low Heat Flow Can and Cannot Tell Us
The reflection profile in Yaeyama also shows a localized structure that could be an intrusion in the shallow subsurface. An intrusion forms when magma pushes into the surrounding rock, but confirming the origin of this structure will require further work such as higher-resolution surveys. AIST's announcement also adds a caveat, saying that local magmatic activity is possible.
The observations of heat flow, reflection profiles, and magnetism each support certain interpretations while leaving others unresolved. Lining up the three kinds of information obtained from the same study area, following the results and discussion of Otsubo and colleagues' paper, clarifies the role each piece of evidence plays.
| Observation | Interpretation it supports | Remaining limitation |
|---|---|---|
| Heat flow generally 40–63 mW/m²; about 730 mW/m² at HPC26 | Heat flow in the shallow part is generally small, though locally hot areas exist | Not a direct measurement of the amount of magma supplied at depth |
| Normal faults spread across both grabens, and a localized structure in Yaeyama | Many faults accommodate the deformation, and a local intrusion may exist | Based on interpretation of two representative cross-sections; deep structure and the origin of the intrusion remain undetermined |
| Magnetic anomalies that vary by location | Consistent with locally magnetized rock bodies rather than widespread shallow volcanic edifices | Magnetism alone cannot uniquely determine the shape or volume of the rock bodies |
This combination supports the interpretation that, under conditions where the magmatic contribution is relatively limited, many normal faults take on the work of crustal extension. It does not establish that coolness caused the faulting, nor does it mean magma is entirely absent underground.
Seafloor heat flow is also affected by conditions other than heat sources below. Thick sediment accumulation, hydrothermal circulation, and transient thermal states may prevent deep activity from appearing directly in surface values. The original paper likewise says that low heat flow is consistent with an explanation of limited magmatic contribution but is not evidence that uniquely demonstrates it.
This survey offers material for examining early rift formation that cannot be fully captured by explanations involving vigorous magmatism alone. However, the dip and offset of faults seen in two-dimensional reflection sections can differ from the actual three-dimensional structure, and the deep crust remains unexplored. While the low heat flow is consistent with earlier observations, neither the original paper nor AIST's announcement presents results of an independent replication of this integrated interpretation.
Next Observations to Connect with Earthquake Risk Assessment
The thermal environment of the crust is also related to the thickness of the layer that can rupture in earthquakes. The research team applied the new heat flow values to an existing assessment framework that links geothermal gradient to the thickness of that layer, and argued that an M7-class earthquake could occur in the southern Okinawa Trough. However, the original paper states explicitly that this estimate carries large uncertainty.
The relationship between heat flow and earthquake magnitude was not experimentally proven as a causal link in this survey. Considering earthquake size also involves how far faults connect, the mechanical properties of the rock, and how stress accumulates. The results do not yield the timing or probability of earthquakes.
According to AIST, the next step is a cruise using the deep-sea drilling vessel Chikyu to investigate crustal heat flow down to about 100 m below the seafloor. In the future International Ocean Drilling Programme (IODP3), the aim is to directly confirm the nature of intrusions and magmatic activity. If heat flow at the shallow seafloor can be compared with deeper heat and rock obtained by drilling, it should become possible to test more concretely how faults and magma share early crustal deformation.