After a major earthquake, our planet keeps vibrating. It “rings” with slow waves that carry clues about what lies far below our feet.
A new study turns that ringing into the first 3D map of how the entire mantle absorbs seismic energy. The big surprise is in the deep mantle, where the places that weaken waves are not the same places where waves slow down.
Listening to Earth’s ringing
The research team, led by Sujania Talavera-Soza with Laura Cobden, Ulrich H. Faul, and Arwen Deuss, analyzed 104 large earthquakes from 1975 to 2018. The work spanned Utrecht University, the Massachusetts Institute of Technology, and Vassar College, focusing on “normal modes,” the planet-wide vibrations triggered by big quakes.
By tracking how those modes split and fade, the scientists built a global model called QS4L3. It focuses on continent-scale patterns, reaching from the upper mantle into the lower mantle.
Until now, global 3D maps of attenuation had mainly covered the upper mantle. The paper was published online on January 22, 2025, at doi.org/10.1038/s41586-024-08322-y.
What “attenuation” means in plain language
Attenuation is a measure of how seismic waves lose energy as they move through rock. Think of it like sound in a room, where a carpeted floor “soaks up” noise more than a tiled one, even if both rooms are the same size.
Higher attenuation means vibrations fade faster, which often points to hotter, softer rock or small pockets of melted material. It can also be sensitive to grain size, meaning how big the mineral crystals are inside the mantle.
That extra sensitivity matters because wave speed alone can be confusing. Temperature and composition can sometimes change speeds in similar ways, but attenuation is far less influenced by bulk composition.
A familiar signal in the upper mantle
In the upper mantle, the new map lines up with what many geophysicists expected. Regions with higher attenuation tend to match areas where seismic waves travel more slowly — a pairing that is usually linked to higher temperatures.
The clearest examples appear beneath mid-ocean ridges, where new seafloor forms. Those ridge zones sit above hotter mantle, so the idea that heat makes waves slower and more quickly damped still holds up.
Not everything is perfectly neat. In some back-arc regions near subduction zones, the model shows high attenuation alongside relatively fast speeds, and the authors say finer-detail maps are needed to explain that mix.
The deep mantle flips the pattern
The biggest twist shows up in the lower mantle. The map finds the highest attenuation in a broad “ring around the Pacific,” even though that same region looks seismically fast rather than slow.
On the other hand, two huge low-speed regions beneath Africa and the Pacific show the opposite behavior. These are large low-seismic-velocity provinces, often shortened to LLSVPs, and they outline roughly a quarter of the region where the mantle meets the core about 1,800 miles down.
In QS4L3, those LLSVP areas come out as comparatively low attenuation, with patterns extending about 750 miles above the core-mantle boundary.
Many geophysicists view that Pacific ring as a “graveyard” of subducted slabs, pieces of ocean plates that sink back into the mantle. Subduction happens when one plate sinks beneath another, and the model hints those slabs stay cold and fine-grained.
Grain size may be the missing ingredient
So how can a region be fast yet highly attenuating, or slow yet weakly attenuating? The team compared their map to predictions from a laboratory-based model of how mantle minerals deform and dissipate energy, paired with mineral physics calculations produced with Perple_X at perplex.ethz.ch.
Those comparisons suggest that temperature is only part of the story in the lower mantle. To get fast speeds together with strong damping, the rock likely has to be relatively cold but fine-grained, while hot rock with larger grains can be slower yet less attenuating.
Grain size sounds like a small detail, but in the mantle it can vary by orders of magnitude. The study’s interpretation points to differences up to about a thousandfold, from crystal sizes around a millimeter, about 0.04 inches, down to scales closer to dust.
Why stable deep “anchors” matter
LLSVPs have been debated for decades because scientists want to know whether they are mostly hot material, chemically distinct material, or some mix of the two. A key point here is that attenuation responds strongly to temperature and grain size, but for the most part it does not respond much to changes in overall chemical composition.
When the researchers translated their temperature and grain-size picture into viscosity — the mantle’s resistance to slow flow — the LLSVPs looked unusually stiff. That stiffness helps explain why these provinces could remain stable for very long periods, acting like deep “anchors” that shape how mantle convection organizes itself.
At the surface, mantle convection shows up as plate motion, earthquakes, and volcanism. This study is not about forecasting the next quake, but it does offer a clearer framework for thinking about the deep structures that guide Earth’s geological engine.
What the new map cannot yet see
The study also highlights limits that are easy to miss in the headline. Because the model targets very long wavelengths, it cannot resolve small features, and it may miss a distinct layer only a few hundred miles thick at the very bottom of the mantle.
There is also uncertainty about exactly where, with depth, the LLSVPs transition into their low-attenuation signature. The authors note that thick structures can hide thinner ones, so some earlier lower-mantle results could still fit within the remaining blur.
Supporting materials are available, including model outputs at GitHub and an archive at Zenodo that includes model coefficients and tools for reproducing the maps.
The main study has been published in Nature.
