SHATTERED HOUSES: Some houses in Lushan County, Sichuan Province, have been torn apart by the earthquake on April 20 (LI GANG)

Geng's prediction was based on his study of the 69 earthquakes measuring 6.0 on the Richter scale and above that took place in north China and the Bohai Sea areas over the previous 2,202 years of history.

He found that between one and 3.5 years before 67 (97.1 percent) of the 69 earthquakes, there were severe droughts in the region.

In the three or four years after 1972, earthquakes rocked the first two of the three areas Geng mentioned. They were the 7.3-magnitude Haicheng earthquake in 1975 and the 7.8-magnitude Tangshan earthquake in 1976 that killed about 240,000 people.

Months before the Haicheng earthquake, abnormal animal behavior had been observed. It was reported that in December 1974 in southern Liaoning Province, snakes awoke from hibernation and froze to death after leaving their caves. Butterflies and ants, which could not be seen in the area in winter, also appeared on streets. Water in some villages' wells surged and ebbed abnormally.

On February 3, 1975, one day before the major shock hit the area, Haicheng's local seismographic observatory recorded frequent foreshocks and reported abnormalities to the provincial government.

On the morning of February 4, the Liaoning Provincial Government issued an earthquake warning, and ordered residents to evacuate.

At 4 p.m. that day, Jiang Chentian, a staff member at the seismographic observatory produced a written report, predicting that an earthquake above 5.0-magnitude would strike the area between 7 and 8 p.m. Jiang made the prediction based on his experience in observing instruments measuring telluric electricity. He found an association between abnormal telluric currents and earthquakes.

At 7:36 p.m., the 7.3-magnitude earthquake struck Haicheng, affecting 8.3 million people in the densely populated industrial area and causing 1,328 deaths. According to Beijing-based Outlook Weekly magazine, without prediction and evacuation, the earthquake would have left a death toll of at least 100,000.

Although the evacuation saved lives, experts disagree on the validity of the prediction.

Li Ping, an academician with the Chinese Academy of Engineering Sciences and a research fellow with the CEA's Institute of Geology, said that some phenomena observed before the Haicheng earthquake were not recorded in other earthquakes. For instance, he said, the frequent foreshocks preceding the Haicheng earthquake did not occur before the Tangshan earthquake.

In addition, Li said that the Haicheng earthquake was predicted on the basis of earth deformation measurement, yet research found that measurements fell within the margin of error and hence could not serve as grounds for earthquake prediction.

The lack of precision presents a dilemma of whether such predictions should be released to the public.

There are two fundamental questions about earthquake prediction, namely, whether earthquakes can be predicted and whether predictions should be reported, said Mei Shirong, founding Director of the former Earthquake Analysis and Prediction Center, which is now known as the Institute of Earthquake Science, under the CEA. Mei has been blamed by some people for failing to predict the devastating Tangshan earthquake.

Earthquake prediction can cause public panic and even turmoil. Zhang Zhaocheng, a research fellow with the CEA's Institute of Earthquake Science said that three months before the two 7.2-magnitude earthquakes that struck Sichuan Province's Songpan and Pingwu counties in August 1976, the Sichuan Seismological Bureau released an earthquake prediction based on a large number of earthquake indicators. Local social order was seriously disrupted. Sixty-one people in a village even committed suicide by drowning, according to Guangdong-based Southcn.com.

Alarm system

Given the difficulty of accurately predicting imminent individual earthquakes, seismologists are seeking alternatives to reduce the potential damage of future earthquakes, including developing earthquake alarm systems.

The Institute of Care-life, a non-governmental research institution in Chengdu, Sichuan, is known for developing an earthquake alarm system.

After the 7.0-magnitude earthquake hit Lushan County on April 20, the system produced alerts via local television networks, the Internet and school broadcasting terminals to neighboring areas, giving Ya'an City 5 seconds of warning and Chengdu City 28 seconds, said Wang Dun, the institute's Director.

Wang said that the system takes advantage of the time lag between non-destructive P waves and destructive S waves emitted from the hypocenter during an earthquake. Since P waves travel faster than S waves, after sensors detect P waves, warning can be sent out before S waves reach the surface.

The system is expected to give people time to evacuate or find shelter, and time to shut off power and gas supply systems and public transit systems.

The early warning system of the Institute of Care-life was officially put into operation in May 2012. It has covered 400,000 square km of land in eight provinces in southwest and northwest China.

Wang said that theoretical research demonstrates that 3 seconds of warning time can reduce casualties by 14 percent, and 20 seconds of warning time can reduce casualties by 63 percent.

Email us at: wanghairong@bjreview.com