TOKYO -- Japan is racing against the clock to prepare for an anticipated major earthquake in Tokyo and to take damage mitigation measures by making full use of available data and artificial intelligence.
Moves by the public and private sectors come as the national government estimates that there is a 70% chance of a magnitude 7-class earthquake hitting directly beneath the Tokyo metropolitan area within 30 years.
But many questions remain unanswered about seismic activity in Greater Tokyo because of its complex subsurface structure.
Naoshi Hirata, chairman of Japan's Earthquake Research Committee, speaks at a meeting in Tokyo on June 23.
In the worst-case scenario, a quake could strike directly beneath the Japanese capital in the evening in winter and be centered in the southern part of central Tokyo. In this scenario, up to 23,000 people could die and 610,000 buildings could be completely destroyed or burn down, according to Japanese government estimates. Economic damage could total 95 trillion yen ($853 billion).
"There is an extremely high risk of an earthquake hitting directly beneath the Tokyo metropolitan area within the next 20 to 30 years," warned Haruo Hayashi, president of the National Research Institute for Earth Science and Disaster Resilience, or NIED.
Hayashi made the remarks at a meeting of government officials, scholars and businesspeople in Tokyo on June 23.
Bracing for nature
A new organization combining the forces of industry, government and academia was created on June 23. The Data Use and Application Council for Resilience was inaugurated to promote efforts to minimize damage from a huge temblor by taking advantage of various data.
NIED played a key role in the establishment of the new organization, whose goal is to build a seismographic network that outperforms the existing one.
In the Tokyo metropolitan area, the earth can shake differently in one location than another that is only a short distance away for several reasons, including the makeup of the area's ground conditions.
The Japanese government and other groups have so far established observation points at about 500 locations in the Tokyo area. But the network is not sufficient to assess shaking and damage in detail.
The council plans to build a wider and more effective seismographic network with the help of the private sector. Participating companies include NTT Docomo and East Japan Railway, known as JR East, as well as Tokyo Gas, which has seismometers at some 4,000 locations.
Artificial intelligence will also be used to analyze earthquake-related data collected by the companies. The council hopes to detect shaking at intervals of only several dozen meters and quickly identify locations that are likely to be affected.
"Our target is to determine how each building shakes," said Naoshi Hirata, a professor at the University of Tokyo and chairman of the government's Earthquake Research Committee. Hirata is doubling as the supervisor of the new council.
Ascertaining how individual buildings sway will make it possible to respond to a powerful quake by giving priority to rescue operations in severely affected buildings and areas. The council intends to establish a system to utilize the data within five years.
Lessons learned
Japan is an earthquake-prone nation, and seismic activity occurs regularly, even if it is not strong enough to be felt, because of its tectonic makeup.
The Philippine Sea Plate and the Pacific Plate are sliding under two inland plates, the Eurasian and the North American plates, triggering various types of earthquakes. Studies have found that many magnitude 7-class quakes have occurred along the fault line of the Philippine Sea Plate.
Knowledge about past earthquakes is crucial for estimating the extent of damage of future ones and taking countermeasures. Researchers are paying attention to one quake in particular that struck the city of Edo, present day Tokyo, in 1855.
It was a magnitude 7-class quake centered directly beneath the city, which was the seat of the Tokugawa shogunate. The death toll from it is said to have reached 10,000.
"It is an earthquake from which we should learn the most lessons," Hirata said.
But it is unknown exactly how the 1855 temblor occurred. One theory posits that the focus was close to the land surface, while another holds that the focus was at a depth of around 100km.
Hirata and other researchers have concluded that the 1855 earthquake "is thought to have occurred either inside the Philippine Sea Plate or on the upper surface of the oceanic plate."
They made this determination after studying 3-D structure data from deep underground and conducting a review of old quake-related records.
According to Toshimi Satoh, a senior researcher at the Shimizu construction company, the 1855 earthquake resembles one that hit the northwestern part of Chiba Prefecture, just east of Tokyo, in 2005 in terms of seismic intensity distribution.
Satoh has estimated that the 1855 earthquake's focus was inside the Philippine Sea Plate at a depth of about 60km under the northwestern part of Chiba. Satoh has also determined the characteristics of the shaking caused by the 1855 quake.
The 2005 earthquake in the northwestern part of Chiba registered an upper 5 on the Japanese seismic intensity scale of 7. It prompted 64,000 elevators in the Tokyo area to be stopped quickly, leaving many people trapped inside.
Shimizu now plans to reflect Satoh's findings in its earthquake-resistance design work for skyscrapers and other buildings.
Magnitude 7-class earthquakes caused by the sinking plates have happened at intervals of about 27 years on average. This year will mark 30 years since the latest such quake, which registered a magnitude of 6.7 and struck off the east coast of Chiba in December 1987.
An anticipated earthquake in Greater Tokyo may not necessarily cause as much damage as the 1855 temblor. But the next one coming is a question of when, not if. In the meantime, Japan needs to conduct further research on the mechanisms of quakes and determine effective countermeasures to minimize the loss of life, injury and damage caused by them.
