“Emergency earthquake warning: seismic intensity lower-six. Twelve seconds, 11 seconds . . .” As speakers loudly relay the warning through the house, Venetian blinds rise in the living room, a gas gas stove switches itself off, the front door is unlocked with a sharp click and automatically propped open with a lever.
This show-house in a northern Tokyo suburb is one of several ongoing trials of earthquake warning technology in Japan. Set up by JEITA, the Japan Electronics and Information Technology Association, it uses information received over the Internet from the Japan Meteorological Agency (JMA) and their nationwide network of earthquake sensors. JMA’s computers analyze data on the first swift-traveling tremors that arrive from the epicenter of a major earthquake to predict where and with what strength the earthquake will strike.
The system can then produce a warning of a few seconds to as long as half a minute, which should be enough time to take minimum precautions to prevent serious injury, says Shinya Tsukada of the Seismological and Volcanological Department of JMA. “You can’t pack up your belongings and run away, but at least you might be able to get under the table.” In March, a JMA-hosted study group of researchers, business representatives and officials from public organizations, issued an interim report regarding the progress of such research and ways that it might be used.
Japan is one of the most earthquake-prone areas of the globe–20 percent of the world’s magnitude 6 and greater earthquakes occur here. Each year, there are more than a thousand earthquakes powerful enough to feel, and major disasters are frighteningly common. In 1995, Japan experienced its most destructive quake of the post-war period, the Great Hanshin-Awaji earthquake, which killed 6,435 people in and around the city of Kobe in central Japan. In October 2004, a large earthquake killed 51 people in the northern Japan prefecture of Niigata.
While virtually no part of Japan is safe from the risk of a major quake, particular attention is focused on the prospect of a major earthquake either in Tokyo, or in the Tokai area west of the capital. A recent government study simulated the consequences of a 7.3 magnitude quake under the north part of Tokyo Bay, a disaster smaller than the 1923 Tokyo earthquake that killed more than 100,000 people, but similar in size to the Kobe earthquake. The study estimated 11,000 deaths and economic damage reaching $955 billion (112 trillion yen) — 850,000 houses were destroyed outright and as many as 7 million people were forced to abandon their homes.
As researchers struggle to produce reliable results from long- and medium-term earthquake prediction systems, recent attention has concentrated on short-term warning systems such as the JMA’s. By using its own network of 200 sensor stations and several hundred set up by other research bodies, the agency can calculate the epicenter of a large quake in as little as two seconds. Although the idea of earthquake warning systems is not particularly new–there are systems in use in California and in Mexico–the system now being developed in Japan promises to be a nationwide network and make use of a range of Internet and mobile communications technology.
Since 2004, the JMA has been transmitting information about imminent earthquakes to select companies, schools and public bodies in a number of trials. A crucial factor is the distance between the epicenter of the earthquake and the area that receives the warning, said Tsukada. “In a really big earthquake, the system may not be useful for people directly above the epicenter, but it may help people further away.”
That principle was clearly demonstrated by two major earthquakes that occurred after the JMA started testing their system. In October 2004, the Niigata area of central Japan was hit by a 6.8-magnitude earthquake. Although the JMA’s system pinpointed the earthquake within seconds, the worst-hit areas were directly above the epicenter, and there was no time to issue a warning. Then, in August of last year, an earthquake hit the northern Japanese prefecture of Miyagi. This time, because the epicenter was 60 miles away in the ocean, the system was able to provide a 16-second warning before tremors reached the heavily populated city of Sendai. The earthquake measured an upper-5 on the Japanese scale of earthquake intensity in Sendai, enough to topple furniture and cause moderate damage to buildings.
“People’s attitudes changed completely after the Miyagi earthquake,” said Tsukada. “They realized that we can really use this technology.” In Tokyo at the time, Tsukada received an automatic cell phone text-message from the system when the earthquake was detected, and then felt the weakened tremors a minute later.
Yukio Fujinawa is the managing director of Real-Time Earthquake Information Consortium (REIC), an NGO that is looking at how information from the JMA’s warning system can be put to practical use. “There are two basic uses for the information,” he said, “one is to stop machines, etc., the other is to warn people.”
Setting up automatic systems to stop assembly lines, halt elevators or alert doctors about to start medical operations, is relatively simple, he said. Reaching a consensus on the second use, how or whether to alert the general public is more difficult. One problem is false alarms; out of 400 alerts since the testing began two years ago, approximately 30 have been mistakes. Another is the possibility of panic. REIC’s Fujinawa suggests a solution could be to introduce the technology gradually. He proposes installing systems in schools and teaching children how to respond to the warnings from a young age. REIC estimates the cost at around $17,000 (2 million yen) for each of Japan’s 55,000 elementary and middle schools.
Warnings might also be issued through the media, much the same way that earthquake reports are broadcast at present. Every year, an automated system passes reports of 200 to 300 earthquakes to the Japanese media, and they are normally broadcast within 2 minutes of the earthquake occurring. Alternatively, Japan’s extensive network of public announcement speaker systems could be put to use. According to REIC, two-thirds of local governments already have suitable systems in place.
One company, the Tokyo start-up 3Soft Ltd., is now developing the world’s first portable home earthquake warning system. They have named it “Digital Catfish,” after the catfish’s legendary tendency to show strange behavior immediately before earthquakes. The PDA-sized receiver picks up wireless earthquake warning transmissions and relays them to smaller speakers placed around a home, or to other safety devices such as those in the JEITA show-house. 3Soft officials hope to price a home-use system at less than $850 (100,000 yen), although they say it is unlikely that the government will allow individual households to buy the product in the near future.
“It’s still not clear whether or not the average household will be able to use our device,” said CEO Hiroyuki Iue, adding that their first customers will probably be businesses or public organizations. Beginning in June, organizations will start applying to the JMA for permission to use the earthquake warning data. 3Soft hopes to start selling the device later this year.
Researchers are also looking at how mobile technology could be used to transmit earthquake warnings. Cell phones already play an important role in the immediate aftermath of earthquakes — though not without problems. “When an earthquake happens cell phones become difficult to use,” said Akira Matsuki, a senior manager at Japanese carrier NTT DoCoMo. “In previous large earthquakes the usage has exceeded capacity by several times to tens of times.” Setting up cell phone capacity to cope with the spikes in usage following a major earthquake or other disaster is prohibitively expensive. Instead, Japan’s major carriers have set up mobile-Internet bulletin board systems to let users leave messages that can be accessed by concerned friends and relatives.
“No matter where you are, if you have a mobile [phone], even though there may be restrictions on voice calls, you still have access to the Internet,” said Matthew Nicholson, Media Relations manager at Vodafone K.K. The systems also link to each other so that messages can be picked up across different carriers. Vodafone’s system has been put into action four times since it was initiated in April last year. Following the Miyagi earthquake in August, 23,000 people checked messages on the site.
A long-term, and somewhat more difficult challenge, is to enable cell phones to relay JMA earthquake warning messages. According to NTT DoCoMo’s Matsuki, developing such a system could both help convey warnings of a coming quake and help cope with the flood of messages afterwards. “We are examining using existing mobile phone technology to send large numbers of messages to users simultaneously,” said Matsuki.
Although he said he couldn’t divulge details of the research, Matsuki noted that a system is unlikely to use present e-mail messages, which take too long to open and read. He also pointed to the need for a standardized system across carriers, and careful consideration of the consequences of the technology. For instance, what would happen if drivers on a freeway decided to stop suddenly when they received an earthquake warning?
The government’s study group on the emergency warning system is due to produce its final report some time between September and December this year. The report is expected to include a road map for implementing earthquake-warning technology. But until consensus is reached on how exactly to use the information, however, it looks like the warning service will only be available to select groups chosen by the authorities. Debate among study group members continues. According to participants, some have raised the question of whether it is fair only to provide warnings of impending catastrophes to those who pay, others have questioned the responsibility of issuing warnings that might be false alarms or could potentially cause mass panic.
“The technology is ready,” said the JMA’s Tsukada. “The argument now is about how to give the information to the public.”
About Tony McNicol
