Cutting edge technologies such as artificial intelligence (AI) and super computers are playing increasingly important roles in China’s efforts to monitor and report earthquake information and assess damage more quickly and accurately.

An automatic AI-driven earthquake monitoring system that can generate seismic source parameters within two seconds is expected to provide new methods for real-time earthquake monitoring.

The system was jointly developed by the research team headed by Zhang Jie, a professor at the University of Science and Technology of China, and China Earthquake Administration (CEA), and put into trial operation at the CEA in December 2018.

Comparisons between 446 earthquake assessment results provided by the system and made using manual computing methods in 2019 have shown that the system provides extremely accurate results.

The system is made up of a set of AI-driven software systems, according to Zhang, who explained that unlike traditional earthquake monitoring systems, it has deep learning ability and is able to quickly handle data on ongoing earthquakes based on millions of seismic materials and seismological theories stored in its memory.

The use of AI technology in the system is similar to having all seismologists working together to handle seismic data, Zhang said.

While traditional earthquake monitoring systems are only able to handle earthquakes greater than magnitude 3, the AI-driven system can provide information on earthquakes of magnitude 1.0 and above, according to Zhang. He calls this an important breakthrough in earthquake monitoring, as small earthquakes are often the forerunners of greater earthquakes.

Accurately locating earthquakes and their fracture mechanisms are usually recognized as the difficult parts of earthquake monitoring. However, the AI-driven system can record and deduce the location and depth of an earthquake according to its seismic waveform, while at the same time calculating its relevant parameters in one or two seconds using its powerful memory database.

These quick results can give early warnings to regions that the seismic wave has not yet reached, winning time for people to take precautionary measures and governments to make rescue plans, Zhang said.

The system is currently the only real-time AI-driven earthquake monitoring system that is running in the world. Its main task is to help the CEA monitor earthquakes in southwest China’s Sichuan and Yunnan provinces.

After one year of trial operations, the system met the relevant technical specifications for conventional earthquake monitoring and early warnings, according to Zhang. However, he added that much time is still needed for the system to be tested and improvements to be made before it can replace operators on duty 24 hours a day.

In addition to AI technology, super computers are also being used in earthquake-related research efforts.

Lin Xuchuan, a researcher with the Institute of Engineering Mechanics, CEA, has developed a simulation system that can assess the preliminary damage done to a city by an earthquake immediately after it happens.

While the traditional way of assessing an earthquake’s damage is to have people survey and collect information on damaged buildings in various areas, the simulation system can provide information on the state of various buildings from earthquakes of different magnitudes, said Lin.

Scientific researchers can use the system to simulate the effects on buildings by earthquakes, which can then help provide cities with detailed suggestions for disaster prevention and mitigation planning, such as arrangements for location and number of shelters, fire stations, and hospitals, according to Lin.

With the increase in the amount of data and the simulation system becoming more precise, Lin felt that a more powerful super computer or supercomputing cloud platform was needed.

On one occasion, Lin completed a computing task in just a few minutes using China’s Tianhe-2 super computer that would have been impossible to accomplish with a regular computing platform.

“For disaster relief in an earthquake, every second is precious. And super computers can help win more time,” Lin said, noting that super computers will become an important strategic resource in smart society, and that their application can offer an opportunity for huge progress in earthquake damage simulation.