Researchers claim to have found a correlation between the occurrence of earthquakes in the Izu Island chain and subtle changes in subterranean geoelectricity, a finding that one day might help develop techniques for predicting temblors.
The team, consisting of researchers from institutions including Tokai University and Tokyo Gakugei University, analyzed the relation between small changes in geoelectricity around Kozu Island, located 170 km southwest of Tokyo, and quakes in the vicinity with a magnitude of at least 3.0, based on data gathered between May 1997 and June 2000.
The geoelectric data were collected during this period through about 20 electrodes buried at intervals of between 100 and as much as 3,000 meters around Kozu. The team studied temblors that struck within 20 km of the island, according to a study published online by a prestigious U.S. science journal.
The researchers observed 19 anomalous changes in the strength and movement of geoelectric currents, 11 of which were proceeded by 3.0-magnitude or stronger quakes within 30 days — a 58 percent rate of occurrence, they reported in the Proceedings of the National Academy of Sciences of the United States of America.
“This rate of probability is statistically significant,” said Toshiyasu Nagao, a Tokai University professor who coauthored the study. “There is debate over the existence of precursors to earthquakes, but (this study) indicates that some may exist.”
The researchers said they excluded geoelectric anomalies caused by factors such as lightning strikes and the sun when determining this rate of occurrence, and reported that a total of 23 temblors with a minimum magnitude of 3.0 struck during the period they examined.
They selected Kozu because of its remoteness and distance from any urban environments, which generate a variety of noises that can effect geoelectricity levels.
Their technique is similar to the so-called VAN method, which was developed in Greece to predict earthquakes based on seismic electric signals. However, scientists have mixed views on the VAN method’s effectiveness and purported 60 percent success rate for forecasting temblors.