The Haicheng incident is what gave people hope that earthquakes might be predictable, says Michael, and what prompted the animal behavior studies by the USGS.
It was later discovered, though, that a rare series of small tremors, called foreshocks, occurred before the large quake hit the city.
"It was the foreshock sequence that gave (Chinese officials) the solid prediction," Michael said.
Still, the Chinese have continued to look at animal behavior as an aid to earthquake prediction. They have had several notable successes and also a few false alarms, said Rupert Sheldrake, a biologist and author of the books, Dogs that Know When Their Owners Are Coming Home and The Sense of Being Stared At.
A reproducible connection between animal behavior and earthquakes could be made, he said, but "as the Chinese have discovered, not all earthquakes cause unusual animal behavior while others do. Only through research could we find out why there might be such differences."
Sheldrake did his own study looking at animal reactions before major tremors, including the Northridge, California, quake in 1994, and the Greek and Turkish quakes in 1999.
In all cases, he said, there were reports of peculiar behavior beforehand, including dogs howling in the night mysteriously, caged birds becoming restless, and nervous cats hiding.
Geologists, however, dismiss these kinds of reports, saying it's "the psychological focusing effect," where people remember strange behaviors only after an earthquake or other catastrophe has taken place. If nothing had happened, they contend, people would not have remembered the strange behavior.
Reporting Strange Behavior
Sheldrake disagrees. Comparable patterns of animal behavior prior to earthquakes have been reported independently by people all over the world, he said. "I cannot believe that they could all have made up such similar stories or that they all suffered from tricks of memory."
More research is needed and is long overdue, said Sheldrake, who proposes a special hotline or Web site where people could call or write in if they saw strange behavior in their animals. A computer would then analyze the incoming messages to determine where they originated. A sudden surge of calls or e-mails from a particular region might indicate that a quake was imminent.
The information would be checked to make sure the observations were not caused by other circumstances known to affect the behavior of animals, such as fireworks, or changes in weather. And to avoid issuing false warnings, Sheldrake said, the data would be used in conjunction with other monitoring devices such as seismological measurements.
"Such a project would capture the imagination of millions of people, encourage large-scale public participation and researchand would be fun," he said. "What is holding this research back is not money but dogmatism and narrow-mindedness."
SOURCES AND RELATED WEB SITES