Deep-Focus Earthquakes Explained

Deep-focus earthquakes occur between 300 and 700 km below the surface. They present seismology with a genuine puzzle, because by any straightforward reading of rock mechanics they should not be able to happen at all.

The problem

An earthquake is brittle failure: rock builds up elastic strain, then snaps along a fault surface and slips.

At 500 km depth the confining pressure is on the order of 20 gigapascals and the temperature exceeds 1,000 °C. Under those conditions rock does not snap. It deforms plastically, flowing slowly like extremely stiff putty. The frictional sliding that drives shallow earthquakes is suppressed entirely — the pressure holding the fault surfaces together is simply too great.

And yet deep-focus earthquakes are recorded routinely, some of them reaching magnitude 8.

Part of the answer: subduction

Deep events occur almost exclusively within subducting slabs. A slab of oceanic lithosphere descending into the mantle carries its low temperature with it and warms only slowly, so it remains far colder — and therefore more brittle — than the surrounding mantle at the same depth.

This explains the geographic distribution. Deep-focus earthquakes trace inclined planes of seismicity called Wadati–Benioff zones that map descending slabs directly, and they are absent everywhere slabs are absent.

But temperature alone is not enough. Even a cold slab at 500 km is under pressure sufficient to suppress ordinary frictional failure. Something else must enable rupture.

Two proposed mechanisms

Transformational faulting is the leading explanation for the deepest events. Olivine, the dominant mineral in the mantle, converts to a denser spinel structure at depth. In a cold slab this transformation can be delayed, leaving metastable olivine persisting deeper than it thermodynamically should. When it finally transforms, the abrupt volume reduction can nucleate a rupture that propagates along a thin shear zone — a mechanism sometimes called anticrack faulting.

Dehydration embrittlement, the mechanism thought to dominate at intermediate depths, may extend deeper if hydrous phases survive further into the slab than expected. Its role at the greatest depths is debated, since most water should have been driven off well above 300 km.

Neither mechanism is settled. Deep-focus earthquakes remain an active research area, and laboratory work at these pressures is extraordinarily difficult.

Why they break so little

Deep-focus earthquakes can be very large and still cause almost no damage.

The 2013 Sea of Okhotsk earthquake reached magnitude 8.3 at 609 km depth — an enormous release of energy, greater than most destructive earthquakes in history. It was felt across a vast swathe of Asia, and it damaged essentially nothing.

The reason is distance. Someone at the surface above a 609 km rupture is 609 km from the source. Seismic energy spreads and attenuates over that path, and higher frequencies — the ones that shake buildings hardest — are absorbed preferentially. What arrives is long-period motion spread across an enormous area.

For the same reason, deep events cannot generate tsunamis. Sea-floor deformation from a rupture that far down is smeared across such a wide region that vertical displacement of the water column is negligible.

The 700 km cutoff

Seismicity stops almost entirely below about 700 km, which is one of the sharpest observations in the field.

Two effects combine. Slabs that have descended that far have generally warmed enough that even their cores deform ductilely. And many slabs stall or deform substantially at the mantle transition zone around 660 km, where a major phase change increases the density of the surrounding mantle and impedes further descent.

Whatever the balance between those causes, the result is that the deepest earthquake ever recorded sits at roughly 700 km, and the mantle below that appears aseismic.