Researchers have come a step closer to understanding how and why the earthquake and tsunami that devastated Japan in 2011 were so surprisingly big. Temperature sensors installed in the fault last year now show that friction between the rocks during the quake was an order of magnitude smaller than previously assumed.
The magnitude-9 Tohoku earthquake shocked the research community by setting a record for the greatest amount of slip ever seen in a fault: some 40–80 metres. No one could explain how or why this happened. In late 2011, a group of researchers mounted a ‘rapid response’ effort to investigate (see ‘Drilling ship to probe Japanese quake zone’).
In the spring of 2012, they managed to install a suite of 55 temperature sensors more than 850 metres into the fault, which itself lies under 6,900 metres of water. Creating an observatory at those depths was in itself a record-breaking achievement. The project faced many challenges: bad weather delayed the installation, shifts in the fault could have crushed the instruments and an earthquake in December could have buried the observatory with landslides. But the team managed to retrieve their sensors on 26 April.
“Amazingly, it seems like the experiment might have actually worked,” says team member Emily Brodsky of the University of California, Santa Cruz. She and a colleague presented their preliminary results at the Japan Geoscience Union Meeting on 19 May.
The temperature measures show how heat dissipated from the fault over time, enabling the researchers to extrapolate back to the moment of the earthquake and to see how much frictional heat was generated during the shift. From this they calculated the coefficient of friction for the fault, and found it to be an order of magnitude lower than the conventional value that has been used since the 1970s. That lower number means less friction.
The result supports the theory that the friction during an earthquake can be dramatically different from the friction during quiet times, perhaps because water in clays is heated by a quake’s shaking, then expands and jacks open the fault. Brodsky says that there are hints that this finding could be generalized to other faults.
The result is consistent with experiments being conducted by Brodsky’s collaborator Kohtaro Ujiie of the University of Tsukuba, who has been trying to recreate the pressure and temperature conditions of this fault in the lab. Both groups hope to publish their results soon.