INSIGNIFICANT ISOTROPIC COMPONENT IN THE MOMENT TENSOR OF DEEP EARTHQUAKES

THE mechanism responsible for deep-focus earthquakes, which occur at depths of 300-700 km in subducting slabs, has been a long-standing problem in geophysics. Unlike shallow earthquakes, deep earthquakes cannot be attributed to frictional instabilities across a fault plane, because of high frictional resistance to sliding at depth. A volumetric change associated with a phase transition, expected to occur at depth 1-3, is often invoked as the physical mechanism; if so, the resulting source mechanism should contain a major isotropic component. Although many researchers have attempted to observe such an isotropic component 4-10, no one has yet convincingly proved or disproved its presence. There exists a portion of the seismogram which is well suited to resolve the isotropic component of deep earthquakes but which has not been analysed by previous workers. Here I use this component in a systematic analysis of 19 large deep earthquakes, and show that no significant isotropic component (< 10% of the seismic moment) exists. A sudden implosive phase change can thus be ruled out as the primary physical mechanism for deep earthquakes.