Phase transformations, earthquakes and the descending lithosphere

Abstract

The energy dissipated from metastable phase transformations is investigated in this paper. In addition to the kinetic energy envisaged by Randall, at least two other major forms of energy (latent-heat and metastability energy) are involved in a metastable phase transformation. The seismic-energy density for metastable phase transformation is of the order of 10 24–10 25 erg km −3 for most mantle silicates. As an example, an earthquake of magnitude M = 7.5 would require only 0.01, 0.03 or 0.04 km 3 of basalt, olivine or pyroxene, respectively, undergoing metastable phase transformation. “Earthquakes” caused by both faulting and phase-transformation mechanisms have been observed in high-pressure laboratory experiments. The energy density associated with a faulting mechanism appears to be four to five orders of magnitude less than that associated with metastable phase transformations of mantle silicates, confirming theoretical expectations. According to all of the thermal models for the descending lithosphere available in the literature, metastable phase transformations almost certainly occur in the descending lithosphere. Enormous energies would be released suddenly in association with these metastable phase transformations. Thus, earthquakes within the descending lithosphere at depths greater than 100 km may be induced by metastable phase transformations. However, metastable phase transformations in the descending lithosphere have little or an opposing effect on the tectonic engine driving the motion of plates into the mantle.