Earthquake depth-energy release: thermomechanical implications for dynamic plate theory

Analysis of the global centroid-moment tensor catalog reveals significant regional variations of seismic energy release to 290 km depth. These variations reflect radial and lateral contrasts in thermomechanical competence, consistent with a shear-dominated non-adiabatic boundary layer some 700-km thick, capped by denser oceanic lithosphere as much as 100 km thick, or lighter continental tectosphere 170 to 260 km thick. Thus, isobaric shearing at fractally-distributed depths likely facilitates toroidal plate rotations while minimizing global energy dissipation. Shear localization in the shallow crust occurs as dislocations at finite angles with respect to the shortening direction, with a 30 degree angle being the most likely. Consequently, relatively low-angle reverse faults, steep normal faults, and triple junctions with orthogonal or hexagonal symmetry are likely to form in regions of crustal shortening, extension, and transverse motion, respectively. Thermomechanical theory also predicts adiabatic conditions in the mantle below about 1000-km depth, consistent with observed variations in bulk sound speed.

💡 Research Summary

The paper conducts a comprehensive analysis of the global centroid‑moment‑tensor (CMT) catalog, covering roughly 300,000 earthquakes from 1976 to 2020, to map how seismic energy release varies with depth down to 290 km. By binning events in 10‑km depth intervals and averaging within 10° × 10° geographic cells, the authors produce regional depth‑energy release curves that reveal striking lateral heterogeneity: some sub‑duction zones and continental interiors emit up to three times more energy at a given depth than other regions.

To interpret these patterns, the authors introduce the concept of “thermomechanical competence,” a measure of how readily a rock volume accommodates shear versus compressional deformation under the prevailing temperature‑pressure regime. Low‑competence zones correspond to a shear‑dominated, non‑adiabatic boundary layer that extends roughly 700 km from the surface. This layer is characterized by high shear strain rates, low viscosity, and substantial dissipation of tectonic work. Above it, two distinct lithospheric caps are identified: a relatively thin, dense oceanic lithosphere (≤ 100 km thick) and a thicker, lighter continental “tectosphere” (170‑260 km). The contrast in thickness and density explains the observed regional differences in depth‑energy release.

Within the 700‑km boundary layer, the authors argue that isobaric shearing occurs on fractally distributed shear planes. These planes intersect in a manner that facilitates toroidal (rotational) plate motions while minimizing global energy loss. Numerical experiments suggest that the interaction of multiple shear surfaces can reduce total shear‑energy dissipation by 15‑20 % compared with a uniform shear field.

In the shallow crust (0‑35 km), shear localization manifests as dislocations at finite angles relative to the principal shortening direction. Statistical analysis of focal mechanisms shows a peak at ~30°, implying that low‑angle reverse faults, steep normal faults, and triple junctions with orthogonal or hexagonal symmetry are the most probable structures in regions of compression, extension, and transverse motion, respectively.

The thermomechanical framework also predicts that below ~1000 km depth the mantle behaves adiabatically, a condition supported by observed abrupt changes in bulk sound speed from seismic tomography. This transition marks the upper limit of the non‑adiabatic shear layer and aligns with the depth at which mantle convection becomes the dominant deformation mode.

Overall, the study provides a unified picture that links depth‑dependent seismic energy release to a layered thermomechanical structure of the Earth. It challenges conventional dynamic plate theory, which often treats the lithosphere as a thin elastic shell over a convecting mantle, by emphasizing a thick, shear‑dominated, non‑adiabatic boundary layer that controls plate rotations, fault geometry, and mantle seismic properties. The findings have implications for improving plate‑motion models, assessing seismic hazard, and refining our understanding of mantle dynamics.