Earthquake Prediction

One of the urban myths that I'm consistently confronted with here in China is the idea that animal behavior can be used to predict earthquakes. It came up again in my English class yesterday, and so I thought I'd do a little digging.

One valid way of predicting earthquakes is based on the fact that "P-waves", which are non-destructive, travel faster than "S-waves", which cause most of the damage. They could theoretically provide a warning of 1 second for every 5 miles distance from the epicenter. I.e. it could only provide a warning of a few seconds, up to maybe a minute or so for a large earthquake, depending on how far away you are from the epicenter.

A very famous instance of a successful earthquake prediction happened in Haicheng 海城, China (in Liaoning 辽宁 Province) in February, 1975. Chinese officials ordered the evacuation of Haicheng (population about 1 million) one day before a magnitude 7.3 earthquake struck. Only 2,041 people died and 27,538 were injured. It was estimated that the number of fatalities and injuries would have exceeded 150,000 if no earthquake prediction and evacuation had been made.

Prior to the quake, over a period of months, changes in land elevation and ground water levels were observed. There were widespread accounts of peculiar animal behavior. There was regional increase in seismicity (which later was recognized as foreshocks) which had triggered a low-level alert. Then, the day before the quake, there was a marked increase in foreshock activity, which triggered the evacuation.

However, in July 28, 1976: a magnitude 7.6 earthquake struck the city of Tangshan, a city with approximately one million inhabitants, without warning. None of the precursors observed near Haicheng were observed this time. There were 250,000 fatalities and 164,000 injured.

Earthquake Prediction and Myths summarizes: "The point is, consistency of the precursors has always been a problem in earthquake prediction research."

As far as the animal behavior goes, from what I've read about the Haicheng earthquake (which admittedly is not a lot) it seems possible that the animals were reacting directly to the foreshocks, and that the foreshocks themselves were the primary indicators of the major earthquake. That is to say, if I were setting up an earthquake-prediction center, I would set up scientific equipment to detect foreshocks, as opposed to centers for observing animals, since the former are really what I'd want to monitor.

Yet, I still have a nagging question about the blanket statement by mainstream scientists that "Science cannot predict earthquakes -- and may never have the ability." [also from "Earthquake Prediction and Myths"]. Just because the precursors are not consistent doesn't mean they are totally useless. In fact, the Chinese authorities did successfully predict the Haicheng earthquake, and saved thousands of lives. Whether based on foreshocks or whatever, the prediction scheme is not perfect, but it did yield some results. Any prediction scheme will have some false positives (predicting earthquakes that never happen) and some false negatives (failure to predict earthquakes that do happen) but there are ways to analyze data to determine if the scheme is statistically significant, and if so, worthwhile.

As far as animals predicting quakes goes, though, I am just not buying it. I haven't read any plausible explanation of how it could be possible. A typical web page arguing in favor is Whispers from the Earth, which lists some things to look for, including:
- Some animals eat more
- ... cats, will ... clean themselves frequently or be especially attentive to their owner, crying or acting nervous.
- Dogs ... become preoccupied with sniffing the ground.

These are brilliant! We should all run for cover the next time we see any dogs become preoccupied with sniffing the ground.

On the flip side, reports of animals behaving strangely before an earthquake, that come out after the earthquake has occurred, should be viewed suspiciously (or better yet, thrown out completely), as this short FAQ answer puts it succinctly: "Animals change their behavior for many reasons and given that an earthquake can shake millions of people, it is likely that a few of their pets will, by chance, be acting strangely before an earthquake."