Left to its own devices, shoddy research has a way of eventually being discarded by the internal, self-critical processes of science. But shoddy science, once reported in the news, has a way of staying forever in cyberspace and, therefore, in the public mind. One example is the theory that building dams and reservoirs can cause major earthquakes. This idea has been around for years, but no one has ever established a scientifically creditable link.
It has now resurfaced again in major newspapers around the world. Some earth scientists in China and the United States have apparently linked a four-year-old reservoir - built close to a geological fault line - to the Sichuan earthquake that left 88,000 people dead last year. The quake, they say, could have been triggered by the weight of 320 million tonnes of water in the Zipingpu reservoir less than 1km from a well-known major fault line.
I am as much against dam-building as any critic who has come down on this mania of many developing countries. Even if all the economic benefits authorities have claimed for their mammoth engineering projects are real, the human costs are hardly worth it. The Three Gorges projects, for example, have displaced millions of people.
I was, therefore, initially intrigued by the latest reports about the Zipingpu dam, until a physicist friend disabused me of the notion.
One thing that the most fair-minded geophysicists and earth scientists agree on is that it is basically impossible to predict earthquakes. This unpredictability is, indeed, a basic principle in chaos theory, according to my friend. Now, if those who believe dams can cause quakes are right, they should be able to explain and predict such catastrophes, at least insofar as dams near fault lines are concerned. This would contradict the idea that quakes are chaotic phenomena. But my friend's point is that earthquakes follow a well-established power law in chaos theory. 'There is overwhelming evidence that events satisfying a power law cannot be predicted by measuring known scientific parameters,' he said. 'That's why geoscientists have totally given up hope of using strain gauges along faults to predict earthquakes.'
He believes earthquakes are so-called 'black swan' that which cannot be accurately predicted. Also included among such phenomena are the crashes of stock markets. 'They all fall under the power law,' he said. 'There are no known particular reasons or threshold values of parameters that directly trigger these events.