CPT Q. 011: Why do earthquakes sometimes occur across zones broader than the width of a plate?
Q. 11. QUESTION Why is it that beneath trenches, earthquakes sometimes occur across a much broader region than the width of a plate?
Response: Earthquakes that occur well behind the trench location tend to be deep-focus earthquakes which occur at depths between 300 and 700 km beneath the earth’s surface. Because subducted lithosphere should not exhibit brittle behavior at such depths, the mechanism responsible for these deep earthquakes has stirred controversy since their actual depth was first verified more than 70 years ago. Because mineralogical phase changes occur in the lower part of the upper mantle where these earthquakes are most frequently observed, a possible and leading candidate mechanism has been the catastrophic transformation of metastable olivine into the higher density spinel phase. Because of the low temperatures in the core of the subduction slab, this phase transition likely may not always spontaneously occur as the slab passes through the depth where the phase transition otherwise ought to take place. When this is the case, metastable olivine is transported to greater depths and has the potential to transform rapidly to the spinel phase, provided there is some process to initiate this transformation. However, a simple volumetric implosion of the low density olivine phase to produce the higher density spinel phase does not match the pattern of earthquake waves these earthquakes radiate—a pattern which typically implies a large amount of shear deformation.
However, about 20 years ago H. W. Green and P. C. Burnley in “The failure mechanism for deep-focus earthquakes,” Geological Society, London, Special Publications 54, 133-141, 1990, described the mechanism now generally thought to account for these deep focus earthquakes. In the abstract of this paper they summarize their findings:
Experimental deformation of Mg2GeO4 olivine at pressures between 1 and 2 GPa in the spinel stability field has led to discovery of a faulting instability that develops at the kinetically-controlled threshold of transformation. Very fine-grained olivine and spinel are found in fault zones. Deformation at lower temperatures is ductile; transformation is inhibited and specimens are very strong. Deformation at higher temperatures also is ductile but transformation is rapid and specimens are much weaker. Detailed examination of the microstructures of specimens deformed in the faulting regime lead to an anticrack theory of faulting that explains the experimental data and provides a fundamentally new mechanism for deep-focus earthquakes. The new mechanism is analogous to the Griffith theory of fracture; nucleation and growth of spinel under stress produces spinel-filled microanticracks normal to the maximum compressive stress that link up to produce faulting. The friction paradox for deep earthquakes is resolved because this faulting process provides a fine-grained, superplastic, ‘lubricant’ for faults. The temperature distribution within subducting slabs of lithosphere requires that the conditions of instability are reached as a natural consequence of subduction; metastable olivine in the interior of deep slabs warms to a critical temperature where faulting ensues in the presence of a shear stress.
To summarize, Green and Burnley used the germanium analog mineral, Mg2GeO4, instead of silicate olivine, (Mg,Fe)2SiO4, to investigate the mechanics of this phase transition in the laboratory in a large enough volume to be able to observe and characterize the actual faulting process. The germanium analog is softer and changes to the spinel structure at much lower pressure than the silicate mineral. Their experiment appears to elucidate how this phase transition can unfold extremely rapidly and also generate large-scale shear motions within the core of a subducting slab.
Another observation that points to the likelihood of the mechanism involving the rapid transformation of olivine to spinel and possibly other lower density phases such as pyroxene transform to their higher density phases is that deep focus earthquakes cease abruptly below a depth of about 680-700 km, which represents the boundary between the upper and lower mantle. This is the depth at which the major upper mantle phases are converted to the yet higher density phases perovskite and magnesiowuestite. Hence, whatever the mechanism is, it shuts down when these transitions between upper mantle mineral phases no longer can occur.
These observations and conclusions apply equally to both UPT and CPT.