Use machine learning to reveal patterns of global earthquake ruptures

Li Zefeng, a researcher at the University of Science and Technology of China, used machine learning methods to summarize the source time function characteristics of more than 3,000 large-scale earthquakes (above 5.5 magnitude) around the world, showing the similarity and diversity of global earthquake rupture processes in a panoramic manner, deepening the understanding of seismic energy release patterns, and having enlightening significance for early earthquake early warning. The results of the research were published in the Geophysical Research Letters.

Figure .jpg

Distribution of global seismic source time function in the implicit space of the variational autoencoder (a) and reconstruction of the global seismic rupture pattern manifold (b) Courtesy of the University of Science and Technology of China

Earthquakes are one of the important natural disasters facing human society. In the past 20 years, the world’s largest earthquakes have caused nearly 1 million casualties and countless economic losses. There are many kinds of seismic rupture processes, and objectively measuring their similarities and differences helps to understand the physical processes of earthquakes and the early prediction of earthquake magnitude.

However, previous studies or the average rupture process of multiple earthquakes cannot measure the range of global seismic differences, or statistics based on certain rupture characteristics, and cannot make a systematic comparison of the entire rupture process.

Li Zefeng used the variational autoencoder in deep learning to compress and reconstruct the source time function of more than 3,000 medium and large earthquakes around the world in two dimensions, and comprehensively demonstrated the global seismic moment release mode and quantitative distribution.

The study found that medium and large earthquakes are mainly simple ruptures and less complex ruptures, and reveal the distribution rules of two types of special earthquakes that have been less concerned but are very important in previous studies: energy release concentrated in the escape mode earthquake in the late rupture stage, and complex mode earthquake with multiple energy releases.

Li Zefeng explained, “The escape mode refers to the earthquake releasing very little energy in the initial stage and then evolving into a large earthquake in the later stage, which is the most challenging type in earthquake early warning and it is easy to underestimate its destructiveness.” Complex earthquakes refer to the release of seismic energy in multiple times, while the vast majority of earthquakes are single releases, and complex earthquakes are also a type of earthquake that is easy to underestimate destructiveness. ”

In addition, Li Zefeng found that the energy release mode of large earthquakes has a weak magnitude dependence, that is, super earthquakes tend to start in the form of releasing lower energy. This provides useful implications for the predictability of final magnitudes in earthquake early warning.

In his reviewers, jean-Paul Ampuero, an internationally renowned geophysicist, said, “The study is an interesting, original, and timely work that is important for understanding the potential limitations of earthquake mechanisms and earthquake early warnings.” (Source: China Science Daily Wang Min)

Related paper information:

Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button