SEISMOLOGISTS should give up trying to predict when earthquakes will strike, according to many of the scientists attending a meeting on quake forecasting in London last week. The money lavished on predicting quakes would be better spent on limiting the damage they cause, they argued.
Robert Geller of Tokyo University told the meeting, organised by the Joint Association for Geophysics, that despite a huge international effort spanning many decades, not a single reliable sign for an impending earthquake had been identified. 鈥淎ll the work in this field has consisted of incorrectly attributing significance to signals at or below the [background] noise level,鈥� he said.
It was 鈥渉ighly unlikely鈥� that quake precursors exist, claimed Geller, who accused some researchers of deliberately exaggerating the success of quake forecasting. 鈥淢any geophysicists have succumbed to the temptation of securing funding by labelling their work as earthquake prediction research. This can usually unlock government coffers.鈥� Geller has long been an outspoken critic of earthquake forecasting methods. But his sentiments are shared by a growing number of seismologists. 鈥淒eterministic prediction is just absolutely not possible,鈥� said Stuart Crampin of the University of Edinburgh.
Advertisement
The increasing mood of cynicism follows the recent controversy surrounding the claim by a team at the University of Athens that they had successfully forecast quakes by measuring electrical disturbances. Some seismologists have criticised the Greek researchers for being too vague about precisely when and where a quake will occur, and how severe it will be.
Many researchers now believe that earthquakes are 鈥渃ritical鈥� phenomena鈥攅vents that occur in systems teetering on the brink of massive change. Predicting when a critical phenomenon will arise is almost impossible. This means that earthquake prediction may be as difficult as guessing which snowflake will trigger an avalanche. 鈥淚f earthquake dynamics is so critical, we may be forced to conclude that the reliable prediction of individual earthquakes is inherently impossible,鈥� said Ian Main of the University of Edinburgh.
However, if earthquakes really are critical phenomena, it may help engineers to put codes for the construction of quake-resistant buildings on a firmer scientific footing. Current codes are based on empirical rules of thumb giving the probability of quakes of different intensities. But mathematicians have shown that critical phenomena obey 鈥渓aws鈥� of the type assumed by earthquake engineers when they design buildings.
