After last week's deadly earthquake in southwest China's Sichuan province, quake prediction has become an issue of intense public concern. However, accurate predictions in the short term are indeed "very difficult," said Lucile Jones, a seismologist at the U.S. Geological Survey (USGS),in a recent interview with Xinhua.
When asked what factors affect the prediction of earthquakes, she said: "This depends on what you mean by predict."
On the spatial distribution, that is, indicating the location of likely earthquakes, "we (seismologists) do an excellent job," said Jones.
However, when it comes to the time and magnitude, prediction becomes difficult.
Magnitude is determined only during the earthquake, by measuring how far the rupture that starts at the epicenter moves down the fault, she said.
If the rupture travels 100 meters, it will be about magnitude 4. If it travels 300 km, the quake will be close to magnitude 8, said Jones.
But what stops the earthquake may not be connected to what makes it start and so information about the size may not be discernible before the earthquake begins.
"Short-term (prediction) is difficult unless the earthquake has a foreshock," she said.
Smaller earthquakes occurring near the first one are called aftershocks.
Only about 5 percent of the time is the triggered earthquake bigger than the first one. The first one is then called a foreshock and the triggered event the main shock.
She said after the first quake happens, the probability of another event is much higher than it usually is, and therefore the likelihood of successful prediction of the triggered event is relatively higher.
In California, the USGS has a website that shows the real-time probability of aftershocks all over the whole state, said Jones, who is USGS' Chief Scientist of Multi Hazards Demonstration Project for Southern California.
Compared to the short term, long-term prediction of earthquakes is fairly mature and usually based on fault history. Governments can set up building codes according to the long-term prediction results in specific areas.
In China and other countries, people can sometimes predict earthquakes or other natural disasters when some unusual events such as unusual behavior among animals or strange astronomical phenomena occur, suggested Jones, while clarifying that there is currently no scientific evidence or research to support this.
"Every effort to prove these has shown there is no solid signal," she said.
Maybe the most important action for earthquake prediction would be for governments all over the world to inject more funds into earthquake research, Jones suggested.
"The whole earthquake monitoring and research program in the United States is funded at 50 million U.S. dollars for the whole country -- the same value (not adjusted for inflation) that we had in 1990," said Jones.