Tsunami made worse by 'pop-up' of sediments
Tsunami created by undersea earthquakes can be made much larger by the "pop-up" movement of large amounts of sediment, research suggests.
These quakes release huge amounts of energy as tectonic plates which stick as they pass each other suddenly slip.
But a study in Earth and Planetary Science Letters suggests that wedges of sediment scraped from the plates can pop up, boosting the resulting tsunami.
It suggests that spotting these wedges could improve large tsunami prediction.
Megathrust earthquakes are the sudden release of "elastic" energy stored as tectonic plates grind against each other, deforming but not slipping.
Typically, as in the Japan earthquake of March 2011 that caused a massive tsunami and the Fukushima nuclear disaster, this occurs at a subduction zone, where one plate is sliding beneath the next.
The quake arises when the deformation becomes too great and the stored energy suddenly releases.
The sudden plate movement is transmitted into the ocean, inducing a tsunami.
Now, two eminent seismology researchers from Cambridge University, Dan McKenzie and James Jackson, have come up with a different idea.
Prof McKenzie, a founder of the idea of plate tectonics, explained that the idea arose from studies of the sea floor surrounding the site of the 2011 Tohoku earthquake off Japan.
"A Japanese submersible went down after the Tohoku earthquake and found not at all what everybody expected - which is that actually the tsunami was generated by the Japanese plate overriding the Pacific plate and moving upwards," he told BBC News.
"But what this submersible found was... there was a fault which had moved the other way: instead of Japan moving upwards over the Pacific (plate), actually Japan had moved downward."
What the pair proposes is that the resulting tsunami was made far worse by the wedge-shaped "accretionary prism" at the boundary between the two plates.
The prism is a collection of loose rock and mud that builds up gradually as one plate slips below another and fragments are broken off each.
When the plates finally violently slip, the point of the wedge is squeezed. This fires the material upward and outward, turning the energy of the plate into the energy of the ejected rock.
"Let's say you have something wedge-shaped on the floor and you jump on it, the wedge will shoot sideways, and that's we think happens," Prof McKenzie explained.
"The weight of the stuff on top shoots a wedge of stuff out. Put some water on the floor and do the same thing, and that will produce a wave."
The team went on to study the sites of other unexpectedly large tsunami from the past, including one off Nicaragua in 1992, Sumatra in 2004, and Java in 2006 - finding similar characteristics of the faults.
That is promising, because it potentially provides a means to anticipate where the greatest tsunami risks lie - it is as simple as using submersibles to find faults that also have accretionary prisms.
And because the resulting tsunami is more a product of the size of the prism rather than the size of the earthquake, the mechanism could explain why sometimes comparatively small earthquakes result in huge tsunami.
Prof McKenzie said that a better analysis of which faults could result in the biggest tsunami could overcome the human factors that have stymied some warning systems.
"There's a huge enterprise around the Pacific of predicting them, and the false alarm rate is very high - people evacuate a few times and then nothing happens," he said.
"That's one of the reasons they're so dangerous - every now and again you have some little earthquake like one that didn't even wake people up in Indonesia but they heard the roar of the wave coming in, but the shaking was so small that they hadn't done anything."