Stanford Geophysicists Produce First Global Map of Continental Mantle Earthquakes, Cataloging 459 Deep Tremors Since 1990
A Stanford-led study published in Science presents the first systematic global catalog of earthquakes originating in the continental mantle, identifying 459 events since 1990 and revealing unexpected clustering beneath the Himalayas and the Bering Strait.
Overview
Most earthquakes originate in Earth’s crust, the relatively thin outer shell that extends down roughly 30 to 50 kilometers beneath the continents. But a small and poorly understood fraction occurs deeper, in the mantle layer that lies between the crust and the planet’s molten core. A study published on February 5, 2026, in Science by Stanford geophysicists Shiqi (Axel) Wang and Simon Klemperer now provides the first comprehensive global catalog of these continental mantle earthquakes, identifying 459 confirmed events since 1990 and revealing that they cluster in geologically significant regions.
The work fills a long-standing gap in seismology. Standard earthquake catalogs have historically assigned default depths to events when precise measurements were unavailable, making it difficult to determine whether a given tremor originated in the crust or the mantle. Wang and Klemperer developed a waveform-based method that eliminates this ambiguity, opening a new window into deep-Earth dynamics.
How They Found the Invisible Earthquakes
The technique relies on the contrasting behavior of two types of seismic waves. Sn waves travel along the uppermost mantle, propagating efficiently through what geophysicists call the mantle “lid.” Lg waves, by contrast, are high-frequency vibrations that bounce through the crust and travel poorly, if at all, through the mantle. By comparing the amplitude ratio of these two wave types at seismic monitoring stations, the researchers could determine whether a given earthquake originated above or below the Mohorovicic discontinuity, the boundary separating crust from mantle.
Starting from a dataset of more than 46,000 continental earthquakes recorded by the U.S. Geological Survey since 1990, Wang and Klemperer narrowed the candidates to 10,770 events with high-quality nearby comparison data. From these, they confirmed 459 as genuine continental mantle earthquakes, events that occur roughly 100 times less frequently than their crustal counterparts.
“Our approach is a complete game-changer because now you can actually identify a mantle earthquake purely based on the waveforms,” Wang said.
Where Mantle Earthquakes Cluster
The resulting global map reveals that continental mantle earthquakes are not randomly distributed. Major concentrations appear beneath the Alpine-Himalayan belt, the Lake Baikal region in Russia, the Bering Strait, southwestern China, and portions of the western United States. The Tibetan Plateau presents a notable pattern: mantle earthquakes are common along its edges but conspicuously absent from its interior.
These clustering patterns correlate with regions of active tectonic collision, extension, or deep thermal anomalies, suggesting that mantle earthquakes are not random artifacts but markers of specific geodynamic processes. In some regions, they may indicate zones where crustal material has been recycled into the mantle through subduction, while in others they could reflect convective stresses or thermal weakening of the uppermost mantle.
Why Mantle Earthquakes Matter
The existence of earthquakes in the continental mantle has been a subject of scientific debate for decades. The prevailing view held that mantle rock beneath continents was generally too hot and ductile to fracture brittlely, the mechanism by which earthquakes occur. Klemperer’s explanation challenges this assumption: “Rocks with higher melting points stay solid at hotter temperatures, meaning the mantle can remain strong enough to break.”
Understanding where and why mantle earthquakes happen has practical implications. Because these deep tremors reveal information about the mechanical state of the upper mantle, they provide constraints on the forces that drive tectonic plate motion, generate magma that feeds volcanoes, and control the long-term deformation of continents. The new catalog also raises questions about whether crustal and mantle earthquakes interact, potentially triggering each other in ways that current seismic hazard models do not account for.
What Comes Next
The global catalog opens several research directions. Wang and Klemperer plan to investigate whether continental mantle earthquakes occur in aftershock sequences, which would suggest stress transfer between events, or whether they are isolated phenomena driven by local thermal conditions. The team also intends to examine whether heat-driven convection processes within recycled crustal material play a role in triggering these deep tremors.
The study’s methodology is itself a contribution. Because the Sn-to-Lg ratio technique requires only standard waveform data from existing seismic networks, it can be applied retrospectively to historical records and extended to new monitoring data as it accumulates. This means the catalog of 459 events is likely a lower bound; as the method is refined and applied to additional datasets, the true global population of continental mantle earthquakes may prove substantially larger.
For seismology as a field, the work represents a shift from treating the crust-mantle boundary as a simple dividing line to recognizing it as a dynamic zone where earthquakes on both sides may be mechanically coupled. That recognition could eventually inform how earthquake hazard is assessed in tectonically active regions worldwide.