Robert Geller, an American seismologist teaching in Japan, has been fighting a lonely battle against what he says is a hopeless national initiative for predicting big earthquakes.
“Earthquake probability information announced by the government is quite meaningless,” Geller, 62, said at an open seminar held by an academic society in Kobe, which was heavily damaged by a quake in 1995.
“An earthquake could occur anywhere without warning. You should prepare for the unpredictable,” he said.
While at the California Institute of Technology, better known as Caltech, in the 1970s, Geller studied under Hiroo Kanamori, a noted Japanese seismologist who was teaching there at that time.
Geller later moved to Japan and became the first-ever tenured foreign faculty member of the University of Tokyo in 1984. He is currently a professor in the Department of Earth and Planetary Science at the university’s Graduate School of Science.
When he came to Japan, Geller was baffled by the way the government was tackling earthquake risk.
In 1978, Japan enacted a law to guard against future earthquakes based on the assumption that prediction would be possible. The Meteorological Agency was charged with the daunting task of spotting dangerous signs and issuing alerts so evacuations and other emergency actions can be taken before a temblor strikes.
The focus of fear was the perception that a huge quake now known as “the Big One” was bound to strike the Tokai region sometime in the coming decades. The public has been so haunted by this threat that the dreaded “Tokai Earthquake” has become a household word.
“Although I had some knowledge beforehand, I was surprised to learn about the details,” Geller said of Japan’s policy for dealing with quake risk.
An earthquake occurs when underground rock slips along a fault. There are countless faults, and such slips continually cause tremors, most of which are negligible.
“Whether a quake becomes powerful or not depends on chance,” Geller said. “Moreover, it is not possible to monitor the state of faults deep underground. Predicting earthquakes is very difficult.”
Japan’s quest to forecast earthquakes began in earnest well before the earthquake contingency law of 1978. In 1965, the government launched a national project to detect telltale signs of huge temblors. The Meteorological Agency and universities worked together to create an extensive network of observation posts. The project continued for more than three decades as a succession of five-year plans, without producing any tangible results in accurately forecasting quakes.
The project had its roots in a proposal in the early 1960s jointly written by several experts calling for quake prediction efforts. The paper implied that if its proposals were fully implemented, prediction might become possible within a decade.
But Geller pointed to the paper’s lack of a scientific approach, saying it offers no explanations on how the monitoring data can used to forecast earthquakes.
Speculating on the thinking of the paper’s authors, Geller said that exaggerating the chance of succeeding in earthquake prediction may have been necessary as a ploy to secure a steady flow of government funds for earthquake research.
“But if you tell a lie, you will eventually betray the trust of society,” Geller said.
Japan was not alone in attempting earthquake prediction at the time. According to Geller, expectations for quake prediction mounted in the United States in the 1970s, creating a phenomenon he termed a “prediction research bubble.”
But the bubble collapsed in the 1980s.
In 1991, Geller launched a public broadside against Japan’s persistent obsession with earthquake prediction, contributing an article questioning the nation’s seismology tactics to the British science journal Nature.
In his eyes, Japan’s anti-earthquake policy was still bedeviled by the lack of scientific approach, raising concerns that “lies would snowball if left unchecked.” Although Geller was the most visible critic, there were also skeptics among Japanese scientists.
Teruyuki Kato, 61, chairman of the Seismological Society of Japan and a professor at the University of Tokyo’s Earthquake Research Institute, remembers the high expectations that earthquake prediction was drawing when he was starting to study seismology in graduate school decades ago.
However, Kato was soon confronted with a harsh reality.
“As I proceeded with my research, I realized that prediction would be difficult,” he recalled.
Nevertheless, Kato was asked to report on his progress toward forecasting ability when the government mapped out a new prediction plan.
Fellow seismologist Hideki Shimamura repeatedly criticized the earthquake prediction project while he was teaching at Hokkaido University.
Each time he threw a barb at the project, he was summoned to the head office of the university, which is a national institution, for grilling by a member of its board of trustees, who was a former government bureaucrat.
Shimamura, 72, is now a professor at Musashino Gakuin University.
A bit of a sea change came after the Great Hanshin Earthquake of 1995 wrecked Kobe and the surrounding areas. The disaster shook the nation’s belief in the science of earthquake prediction, because the risk of a powerful quake in the region had failed to show up on the radar screens of any seismologist.
The offshore Great East Japan Earthquake of 2011, which devastated parts of the Tohoku region and set the Fukushima nuclear crisis in motion, came as an additional blow to proponents of earthquake prediction because its overwhelming scale and the fatal tsunami it spawned exceeded all assumptions.
As a result of the reviews triggered by the catastrophes, the government has since shifted emphasis away from prediction, at least on the surface, and renamed the national prediction project “disaster mitigation research.”
Still, the earthquake contingency law that epitomizes what has proven to be a flawed policy is still in force.
The Headquarters for Earthquake Research Promotion, a governmental organization overseeing earthquake research, continues to update Japan’s vague quake probability information.
For example, in April this year, the organization released updated information that put the probability of Tokyo being struck by a 7.0-magnitude earthquake within the next 30 years at around 70 percent.
Geller berates news organizations for feeding the public information on earthquake probability announced by the government without questioning the science behind it.
Calculating earthquake probabilities is an approach based on the theory — and not a rock-solid one at that — that earthquakes of a similar scale occur periodically in the same region.
At long last, there is a move among seismologists to revisit the scientific basis for such forecasting.
Geller has urged the government to abolish the earthquake contingency law and rethink its fight against earthquake risk so it can concentrate instead on “more efficient measures.”
“The negative legacy must not be passed to future generations,” he said.