The Unpredictable Earthquake

The old saw about lightning never striking twice in the same place is about as true for lightning as it is for earthquakes: not very. As lightning repeatedly seeks out sites that are good conductors and elevated above ground level, so do earthquakes return to the same areas time after time. Although lightning strikes occur randomly in time, major earthquakes display a periodicity in their "return time" to the same location.

This is because after a major earthquake has occurred, it requires a certain amount of time for gradual crustal movements to build up enough stress to overcome the strength of the earth's crust where failure will reoccur. This length of time varies from place to place, depending on the relative amount of movement taking place and the nature of the crustal material involved.

After the great Alaska earthquake of 1964, it was commonly (and erroneously) believed that all the strain had been released in the area, and that another large earthquake would not occur for thousands of years. In truth, the return time for earthquakes of magnitude 8 or so in the rupture zone which produced the 1964 earthquake is more on the order of a hundred years. The simple fact is that it is not a question of if it will happen again--it's a matter of when.

Unfortunately, the high hopes that were held ten years ago for an imminent breakthrough in earthquake prediction have not been realized, except possibly in limited areas and with limited success.

The underlying principle of earthquake prediction theory lies in the unlikely finding that rocks actually expand when subjected to stress. This is because minute cracks or pores appear in the rocks' fabric, increasing the volume.

This results in a number of physical manifestations which should be observable with the proper equipment. For one thing, the potential epicentral zone should swell. The celebrated "Palmdale Bulge" in California, where surveyors have noted a substantial uplift of several inches in recent years, has failed to produce an earthquake of note despite repeated warnings.

Expansion of the rock and the opening of its pores produces several other phenomena. The velocity of certain types of seismic waves is reduced. Electrical resistivity increases. The increased flow of ground water should change well levels and increase the emission of the radioactive gas, radon, into the well water.

These are only a sampling of the physical changes which "should" be observed if the theories of earthquake prediction developed since the late 1960s are correct.

In fact, in carefully monitored situations, some of these phenomena have been observed, but rarely, if ever, all at the same place and at the same time to warrant the claim of a true prediction.

The most celebrated case of an earthquake prediction was performed by the Chinese when they evacuated the city of Haicheng on February 4, 1975. Following the evacuation, an earthquake of magnitude 7.3 leveled the city. Their prediction was based partially on abnormal animal behavior, a field which western scientists have mostly scorned and only recently begun to take seriously.

However, a year later on July 27, 1976, an earthquake of magnitude 7.6 destroyed Tangshan, also in northeast China, and no warning was given at all. This earthquake killed 655,237 people, making it the second most costly in recorded history (the greatest toll was taken in Shensi, China in 1556 when nearly 1,000,000 people were killed.)

Clearly, earthquake prediction has a long way to go to claim respectability. But then, so does weather prediction.