Earthquakes could be better predicted

Heavy earthquakes occur again and again around the Pacific. Like that off the Japanese coast in March 2011, which in conjunction with the following tsunami around 20 000 People lost their lives and triggered the nuclear disaster in Fukushima. It came as a surprise, if the population had been warned, the consequences would have been less severe, says Onno Oncken from the GFZ German Research Center for Geosciences in Potsdam.

A precise prediction is true at what time and with what force the earth trembles, according to current knowledge impossible. “But the information that the likelihood of earthquakes will increase significantly over the next few weeks will help,” says Oncken. In his opinion, this is exactly what could be possible in the future – using a process that GFZ scientists and international colleagues describe in the journal “Nature”. It would be a huge step forward, but the reactions from the experts are subdued.

A surprising pattern in the earth's crust

The team around lead author Jonathan Bedford has analyzes how the earth's crust before said Tohoku-Oki earthquake 2011 (magnitude 9.0) and before the Maule quake 2010 (Chile, magnitude 8.8). They used data from satellite navigation systems such as GPS and discovered a surprising pattern.

Both locations are located in a subduction zone, where one tectonic plate submerges under the other. Normally the “emerging mainland” – here Japan or Chile – moves away from the subduction zone because the continental plate is compressed a little and thus shortened. In these cases, however, there was a sudden change of direction, the country moved a few millimeters towards the open ocean for a few months before returning to the usual movement. Shortly afterwards there were strong earthquakes.

Drainage of the oceanic plate

The researchers propose the following explanation for the change of direction: The submerging oceanic plate experiences in the deep rock transformations, in which the material becomes denser and thus pulls down stronger and faster. This conversion is secured, but according to the model it happens very quickly and extensively.

In addition, there are drains of the oceanic plate, which act like lubricants and thus make it easier for both plates to slide past each other in depth. This causes the upper continental plate to slide forward, which increases the tension even a few kilometers deep, where the two plates are rigid and interlocked, and soon causes a break – an earthquake.

If the proposed model is correct, could geological researchers use the temporary movement change to predict an impending earthquake? This idea is obvious and Onno Oncken says confidently: “We have a certain hope that we can use it to recognize the increased likelihood of an earthquake.” Of course, with the vague indication that this can be expected “in the next few weeks and months”.


Videografik: So entstehen Erdbeben

Video graphics: How earthquakes occur

But even this imprecise statement is helpful because people adapt to it, internalize escape routes and hold exercises, says the GFZ scientist. “Most deaths from severe earthquakes are not caused by collapsing buildings, because they are constructed accordingly in the typical regions. Instead, they are follow-up events like the tsunami and fires. ”Long-term precautions are essential to protect human lives.

The model has narrow limits

However, the model has narrow limits. It is only suitable for large earthquakes in subduction zones and, moreover, has so far only worked for the two events mentioned. The striking deformation of the earth plates has not yet been identified in other large earthquakes. With better sensors and a denser measurement network, sensitivity should increase and this will sooner succeed, says Oncken.

Jean-Philippe Avouac from the California Institute of Technology is not involved in the study and is reluctant. “There are just two cases where this was shown,” he says. “These striking movements may also occur without an earthquake occurring afterwards.” The hypothesis would have to be supported by further work. That is also what Roland Burgmann (Berkeley) thinks. “I am not yet convinced that the measured deformations really come from the movements of the submerged plate at great depth and that they can above all be used as a sign of two strong earthquakes.”

Should further studies support the assumption , that would certainly be “a big deal”, Bürgmann continues. He believes it is premature to speculate that this model will help to identify the danger of an early earthquake.