On September 25, 2003, a magnitude (M) 8.3 offshore earthquake struck Hokkaido, Japan. The earthquake and ensuing tsunami injured 589 people and resulted in significant damage to port and coastal communities. In 1952, a similar earthquake in terms of epicentral location and magnitude (M 8.1) occurred along this part of the Kuril subduction zone.
These two earthquakes are referred to as the 1952 and 2003 Tokachi-Oki earthquakes. Their similarity allows scientists to determine:
During 2001 and 2002, the U.S. Geological Survey (USGS)'s Coastal and Marine Geology (CMG) team in Menlo Park, CA, hosted visiting scientist Kenji Hirata from the Japan Marine Science and Technology Center (JAMSTEC).
During his time at the USGS, Kenji worked with scientists from both the CMG team and the Earthquake Hazards team to examine the rupture mechanics of the 1952 Tokachi-Oki earthquake, as determined from tide-gauge records of the tsunami.
Results of this study were published earlier this year in the April issue of the Journal of Geophysical Research and presented at the 2003 Annual Meeting of the Seismological Society of America in San Juan, Puerto Rico.
The scientists determined the slip distribution of this event (typically, sections of a fault slip different amounts during an earthquake), as well as what factors might control rupture along a subduction-zone fault.
Preliminary results indicate that the 2003 Tokachi-Oki earthquake occurred primarily in a region that had high slip in 1952 and ruptured in an M 7.0 earthquake in 1962.
This anomalous part of the fault may represent a "sticking point" or asperity. A seamount chain on the Pacific plate is being subducted in this region; however, the spatial relationship between subducted seamounts and the asperity is unclear.
Future research will better define the relationship between high-slip regions from the past earthquakes ( defined primarily by analysis of tsunami records) and the rupture pattern of the 2003 Tokachi-Oki earthquake.
Kenji also worked with CMG scientists during his time at the USGS to analyze micro-tsunamis recorded by one of JAMSTEC's real-time, cabled, deep-sea observatories.
These observatories include two ocean-bottom seismometers (OBSes) and tsunami sensors deployed at various water depths, as well as an "Advanced Technology Observation Station" that includes a current meter, heat-flow meter, hydrophone, conductivity-temperature-depth (CTD) sensor, video camera, and acoustic Doppler current profiler.
The use of micro-tsunamis to determine the source parameters of moderate offshore earthquakes was published this year in Earth and Planetary Science Letters (v. 208). The Hokkaido deep-sea observatory recorded the 2003 Tokachi-Oki earthquake and tsunami. These data, in addition to data from a rapid-response OBS deployment, should provide important information with which to increase our understanding of tsunami generation and the physics of subduction-zone earthquakes.
Even though the 1952 and 2003 Tokachi-Oki earthquakes were very similar in terms of their locations and magnitudes, the tsunami damage patterns were very different.
This result is consistent with the results of modeling by CMG scientist Eric Geist, who showed that local tsunami-runup patterns depend on the details of complex earthquake rupture: namely, the slip-distribution pattern.
Part of the increased runup during the 1952 tsunami may also have to do with the unique hydrodynamics of runup on snow- and ice-covered land. More than 4 m of runup was observed along a stretch of coastline between Kushiro and Monshizu (in northeastern Hokkaido). The 1952 Tokachi-Oki earthquake and tsunami resulted in 28 people dead and 287 injuredthe tsunami alone swept away 91 houses.
In contrast, the tsunami following the 2003 Tokachi-Oki earthquake did much less damage, mainly beaching fishing vessels and sweeping away parked cars. Initial results indicate that most of the slip for the 2003 event occurred at deeper focal depths than that for the 1952 event. In addition, the tsunami that was generated by the 2003 earthquake mostly affected the rugged, sparsely inhabited coastlines of southern Hokkaido.
Fieldwork is currently being conducted to determine the level of ground failure associated with the 2003 Tokachi-Oki earthquake and to test recently developed ground-failure-assessment techniques.
CMG scientist Robert Kayen was invited to participate in an Earthquake Engineering Research Institute (EERI)-sponsored reconnaissance of the 2003 Tokachi-Oki earthquake in Hokkaido to report on the locations and severity of landslides and other damage caused by ground failure.
Extensive ground-failure damage was reported at port facilities, river embankments, and river deposits in southeastern Hokkaido, especially in the Tokachi and Kushiro areas.
In 2001, Rob (then a visiting professor at Kobe University) conducted numerous spectral-analysis-of-surface-wave (SASW) tests at Kushiro and other locations of previous ground failure in Hokkaido and is familiar with the region.
The SASW test, which is a non-invasive way to produce a shear-wave-velocity profile of the ground, is used to quantitatively assess the liquefaction potential of soil. One purpose of the reconnaissance is to evaluate the ground performance at the 2001 SASW test sites during the 2003 Tokachi-Oki earthquake.
Rob will be joined by Yasuo Tanaka (Kobe University), Takahiro Sugano (Japan Port and Harbor Research Institute), Scott Ashford (University of California, San Diego), and David Keefer (USGS, Earthquake Hazards team).
Knowledge about earthquakes and their effects learned in one of the most seismically active regions in the world will be applied to study and help mitigate damage from earthquakes and tsunamis along U.S. coastlines.
in this issue:
2003 Hokkaido Earthquake and Tsunami