Learning from a disastrous megathrust earthquake
Ten years of interdisciplinary studies since the disastrous Tohoku-oki earthquake have improved our knowledge of earthquake-cycle processes and hazard, but prediction of such events remains elusive.
By Naoki Uchida and Roland Bürgmann
On 11 March 2011, a 9 to 9.1 magnitude earthquake occurred off the shore of Tohoku, Japan. This was the biggest recorded earthquake in Japan and one of the five largest earthquakes in the world since the beginning of instrumental observations. It occurred in one of the best monitored areas in the world and has been extensively studied in the past decade. Research results have provided several surprises to the earthquake research community, including the earthquake’s unexpectedly large slip near the trench, the recognition of significant precursory seismic and geodetic anomalies, and the widespread and enduring changes in deformation rates and seismicity across Japan since the event. A recent article published in Reviews of Geophysics gives an overview of a decade of research on the Tohoku-oki earthquake. We asked the authors to explain the significance of this earthquake and lessons learned from it.
What are megathrust earthquakes?
Megathrust earthquakes are plate boundary ruptures that occur on the contact area of two converging tectonic plates in subduction zones. Megathrust ruptures involve thrusting of subducting oceanic plates (here the Pacific plate) under the overlying plates (here Japan as part of the North America or Okhotsk plate). Due to the unstoppable relative motion of the plates, stress accumulates in the areas where the interface of the two plates is locked and is eventually released in megathrust earthquakes.
The world’s greatest earthquakes occur on megathrusts. Megathrust earthquake sources are usually located beneath the sea, which makes it difficult to make detailed observations based on seismic, geodetic, and geologic measurements.
Megathrusts also have the potential to produce devastating tsunamis because of the large ocean bottom vertical movement occurring during the earthquake.
Prior to the Tohoku-oki earthquake, what were the gaps in our understanding of megathrust earthquakes?
Despite many studies of the Japan Trench, there was no consensus on the possibility of magnitude 9 earthquakes before the Tohoku-oki earthquake.
The instrumental records indicated a heterogeneous distribution of up to magnitude 8 earthquakes and repeated slips in the subduction zone. However, the records of the past 100 years did not directly address events with much longer recurrence intervals.
Land-based geodetic observations collected in the decades prior to the mainshock showed strong interplate locking offshore Tohoku. However, the resolution of these measurements was poor in the offshore area, and various ways to compensate the apparent slip deficit, including slow earthquakes, were considered to explain the discrepancies between seismologic and geodetic estimates of megathrust coupling and earthquake potential.
Since the 1980s, geological investigations of coastal tsunami sand deposits provided clear evidence of large tsunamigenic earthquakes that appeared to be substantially larger than instrumentally recorded events. However, the characterization of the ancient tsunami sources and utilization of these results in the evaluation of earthquake hazard was slow.
What exactly happened during the Tohoku-oki earthquake?
The earthquake was a megathrust event, which occurred along the Japan Trench where the Pacific plate thrusts below Japan. The mainshock rupture initiated close to a zone of slow fault slip with foreshocks on the plate interface in the previous months and a magnitude 7.3 foreshock two days prior.
Over the course of about three minutes, the fault slip propagated to fill out the rupture area of roughly 300 by 200 kilometers, catching up a slip deficit that had built up since as long ago as the 869 A.D. Jyogan earthquake. A maximum slip of about 60 meters occurred near the trench, and the resultant tsunami and shaking caused almost 20,000 deaths in Japan.
How has this event improved our understanding of the earthquake cycle and rupture processes?
Thanks to lessons learned from a decade of research, our understanding of the megathrust earthquake cycle and rupture process has improved in many aspects.
Detailed models of the earthquake slip suggest rupture occurred in an area with a large interplate slip deficit indicated by the pre-earthquake geodetic data. Knowledge of the coupling state and complex seismicity near the trench was improved by ocean bottom observations.
Additional geological surveys of tsunami deposits along the coast and observations of landslide deposits (turbidites) on the ocean bottom revealed the recurrence history of great tsunamis and earthquakes. They suggest quite frequent recurrence of tsunamigenic earthquakes that affected the Tohoku area.
The geophysical observations also identified various kinds of possible precursors before the mainshock. Understanding the uniqueness of such phenomena is important to understand the earthquake cycle and may eventually allow for issuing shorter-term earthquake forecasts.
What are ocean bottom observations and how can they improve our earthquake monitoring efforts?
Typical land surface observations of ground shaking and deformation by seismometers, tiltmeters, GPS, InSAR, and any geodetic measurements requiring the transmission of electromagnetic waves or light are difficult or impossible to record at the ocean bottom.
To complement land-based observations, seafloor systems have been developed to monitor the offshore portion of subduction zones. These include (cabled) ocean-bottom seismometers and pressure gauges, GPS-Acoustic measurements (which use sea-surface GPS and sound measurements between the surface and ocean-bottom for estimating seafloor displacements), and ranging using sound waves from ships or between ocean bottom stations.
The ocean bottom measurements better characterize coseismic and postseismic slip, help more accurately monitor the interplate coupling status, locate smaller earthquakes, and observe seismic and tsunami waves much earlier than the instruments on land.
In addition, observations of seafloor sediments provide evidence of ancient and historical great megathrust earthquakes, and boreholes drilled into the megathrust fault zone far offshore allow for examining the fault-zone materials and properties to improve the characterization of structure and fault behavior.
What additional research, data, or modeling is needed to predict future megathrust events more confidently?
Although post-Tohoku-oki studies have better characterized the hazard and a number of possible precursors have been identified, the confident prediction of such events appears impossible in the near future. More detailed investigations of earthquake cycle behavior and interplate locking from the perspective of multiple research fields will further improve the characterization of the conditions of earthquake occurrence and the associated hazard.
A comprehensive compilation of verifiable observations of long-term and short-term precursory processes, including rigorous statistical evaluation of their validity and physical understanding of the processes underlying such phenomena, is important.
While the prospects for reliable short-term prediction of destructive earthquakes may be low, probabilistic operational earthquake forecasting informed by detailed observations of earthquakes and slow-slip activity in the Japan Trench should be possible in the near future.
Why is it essential for earthquake research to be interdisciplinary?
The ability to characterize the nature and hazard of off-Tohoku earthquakes from each disciplinary perspective was limited before the Tohoku-oki earthquake. It appears that it would have been possible to ascertain the occurrence of megathrust events comparable in size to the 2011 Tohoku-oki earthquake if the results from seismic, geodetic, and geological studies had been considered together.
Thanks to a decade of data gathering and research, our understanding of the Japan Trench is much improved compared to what was known before the Tohoku-oki earthquake. However, there are still challenges ahead for each discipline to more fully understand the various facets of megathrust earthquakes and to integrate these findings into a complete picture of the system.