All of Japan is earthquake country, but its most dangerous part is on the Pacific coast of the main island Honshu, just southwest of Tokyo. Here the Philippine Sea plate is moving under the Eurasia plate in an extensive subduction zone. From studying centuries of earthquake records, Japanese geologists have mapped out segments of the subduction zone that seem to rupture regularly and repeatedly. The part southwest of Tokyo, underlying the coast around Suruga Bay, is called the Tokai segment.
Tokai Earthquake History
The Tokai segment last ruptured in 1854, and before that in 1707. Both events were great earthquakes of magnitude 8.4. The segment ruptured in comparable events in 1605 and in 1498. The pattern is pretty stark: a Tokai earthquake has happened about every 110 years, plus or minus 33 years. As of 2012, it has been 158 years and counting.
These facts were put together in the 1970s by Katsuhiko Ishibashi. In 1978 the legislature adopted the Large-Scale Earthquake Countermeasures Act. In 1979 the Tokai segment was declared an "area under intensified measures against earthquake disaster."
Research began into the historic earthquakes and tectonic structure of the Tokai area. Widespread, persistent public education raised awareness about the expected effects of the Tokai Earthquake. Looking back and visualizing forward, we are not trying to predict the Tokai Earthquake at a specific date, but to clearly foresee it before it happens.
Worse than Kobe, Worse than Kanto
Professor Ishibashi is now at the University of Kobe, and perhaps that name rings a bell: Kobe was the site of a devastating quake in 1995 that the Japanese know as the Hanshin-Awaji earthquake. In Kobe alone, 4571 persons died and more than 200,000 were housed in shelters; in total, 6430 people were killed. More than 100,000 houses collapsed. Millions of homes lost water, power or both. Some $150 billion in damage was recorded.
The other benchmark Japanese quake was the Kanto earthquake of 1923. That event killed more than 120,000 people.
The Hanshin-Awaji earthquake was magnitude 7.3. Kanto was 7.9. But at 8.4, the Tokai Earthquake will be substantially larger.
Science Being Done
The seismic community in Japan is monitoring the Tokai segment at depth as well as watching the level of the land above it. Below, researchers map a large patch of the subduction zone where the two sides are locked; this is what will let loose to cause the quake. Above, careful measurements show that the land surface is being dragged down as the lower plate puts strain energy into the upper plate.
Historical studies have capitalized on records of the tsunamis caused by past Tokai earthquakes. New methods allow us to partially reconstruct the causative event from the wave records.
These advances allowed Tsuneji Rikitake to conduct a reassessment (PDF) of the Tokai Earthquake in 1999. Using several different methods, he gauged the quake to have a probability of 35 to 45 percent of occurring before 2010.
The Tokai Earthquake is visualized in scenarios used by emergency planners. They need to create plans for an event that will likely cause about 5800 deaths, 19,000 serious injuries, and nearly 1 million damaged buildings in Shizuoka Prefecture alone. Large areas will be shaken at intensity 7, the highest level in the Japanese intensity scale.
The Japanese Coast Guard recently produced unsettling tsunami animations for the major harbors in the epicentral region.
The Hamaoka nuclear power plant sits where the hardest shaking is foreseen. The operators have begun further strengthening of the structure; based on the same information, popular opposition to the plant has increased. In the aftermath of the 2011 Tohoku earthquake, the plant's very future existence is clouded.
And many municipalities have taken earthquake education seriously, not just in schools but on the Web as well. Some of these pages (such as Tokyo's excellent multilingual PDF) are in English, for the benefit of visitors.
Weaknesses of the Tokai Earthquake Warning System
Most of this activity does good, but some aspects can be criticized. First is its reliance on the simple recurrence model of earthquakes, which is based on studies of the historical record. More desirable would be a physical recurrence model based on understanding the physics of the earthquake cycle, and where the region sits in that cycle, but that is still not well known.
Also, the law set up an alert system that is less robust than it seems. A panel of six senior seismologists is supposed to assess the evidence and tell the authorities to make a public warning announcement when the Tokai Earthquake is imminent within hours or days. All the drills and practices that follow (for instance, freeway traffic is supposed to slow to 20 kph) assume that this process is scientifically sound, but in fact there's no consensus on what evidence actually foreshadows earthquakes. In fact, a previous chairman of this Earthquake Assessment Committee, Kiroo Mogi, resigned his position in 1996 over this and other flaws in the system. He reported its "grave issues" in a 2004 paper in Earth Planets Space.
Maybe a better process will be enacted some day—hopefully long before the Tokai Earthquake of 20xx.