Can Earthquakes Be Predicted

Ian Main

A useful earthquake prediction requires the location, time, and magnitude of a potentially damaging future event be specified in advance, within narrow limits. Further, the probability of occurrence of this prediction (within its stated error bounds) should be shown to be above the background level of random chance in the area of interest. If narrow limits or error bounds cannot be specified, then the resulting uncertainty precludes the practical possibility of an evacuation programme. If the null hypothesis of a random chance event cannot be excluded, then the prediction is not statistically significant. Given the inherently noisy nature of geophysical data, it is also important that data on failures is as easy to obtain as that on successes, including both 'misses' and 'false alarms'. Otherwise the selective publication of data may lead to a spuriously high statistical significance

------------------------------------------------
If any earthquake predictor knows these points stated above yet continues to pursue a very unscientific approach then one may ask: "What is their motive?" Is it to "appear to be successful" while evading the responsibility to prove their point (after perhaps communicating their warning to a lot of people)? Perhaps a precise statement of their purpose will allow a more informed opinion as to how to react to an earthquake warning or prediction from them.

Or maybe they are unable to proceed correctly (though they desire to do so) and need help but do not wish to directly ask for such help?

Of course, the default answer is probably that it's just a lot of work to implement a scientific approach and it is easier to just continue the modis operendi (?) and just keep on pointing out the "maybe's" and "ifs" in the scheme they are using? Only each particular predictor knows for sure (what their motives are)...though others will arrive at their own opinions of them given enough time.

Just some thoughts (.02 worth)

• Can Earthquakes Be Predicted