tokyo

The government committee responsible for overseeing Japan's earthquake prediction programme has approved a report by seismologists that clearly admits for the first time that earthquake prediction is difficult.

In the particular case of the expected Tokai earthquake near Tokyo, on which much of Japan's prediction efforts have focused, the review says that prediction will be possible only in a complex set of circumstances that are unlikely to be met.

The report is being widely interpreted in the Japanese media as admitting that earthquake prediction is virtually impossible. But government agencies involved in the research, as well as leading politicians, seem determined that the prediction programme should continue. The programme employs hundreds of researchers on an annual budget of more than ¥20 billion (US$185 million) and has run for 30 years.

The report was put together by various subcommittees of the Geodetic Council, an advisory body to the Ministry of Education, Science, Sports and Culture made up of astronomers and Earth and space scientists.

These subcommittees were composed primarily of seismologists working on the programme. An external panel of scientists not involved in the research also produced comments on a draft of the report.

The report, which was approved by the full council last week, concludes that predicting the time, location and magnitude of earthquakes is, in general, koonan (“difficult”). The word has connotations of being almost impossible. But its meaning is ambiguous — allowing government agencies and researchers to justify continued research on prediction.

Huge investments have been made in setting up dense observation networks in the Tokai region around the Izu peninsula, as researchers have predicted that a major earthquake can be expected in the near future. A committee of six ‘wise men’ can be called in by the meteorological agency at any time to examine seismic data that might indicate that an earthquake is imminent, and advise the prime minister to issue a warning.

The leaders of the programme have always maintained that prediction of the Tokai earthquake is possible, and this has provided much of the justification for their work. Their confidence was based in part on the belief that the Tonankai earthquake of magnitude 7.9 that occurred in the same region in 1944 showed clear precursors that could have acted as a warning.

But the quality of observation data in 1944 was limited, and much of the 30-year programme has been spent collecting empirical data in the hope of finding clear-cut precursors of earthquakes.

The new review for the first time places severe limitations on the possibility of predicting the Tokai earthquake, however, by stating that, if prediction is to be possible, crustal motions preceding the earthquake must be of the same type and follow the same time dependence and amplitude as the Tonankai earthquake.

In all other circumstances, prediction of the Tokai earthquake will be “difficult”, the review says, adding that no estimate can be given of the likelihood that the required conditions will be met.

For the first time in the history of the programme, the report accepts that there is a “huge gap” between the public's perception of researchers' ability to predict earthquakes and the present level of the science. It calls for the “difficulties” of earthquake prediction to be made “widely known” to the public, which has become disillusioned with the science since the Kobe earthquake of 1995 which caught the public completely by surprise. It also calls for greater efforts in disaster mitigation.

Despite such admissions, the government agencies backing the programme have expressed their determination to carry on with the research. The chief cabinet secretary, Seiroku Kajiyama, has said: “It is a matter of course that the government should continue efforts [to predict earthquakes] if there is a possibility of earthquake prediction in the future”.