Laboratory investigations probing earthquake source process

Abstract

Earthquakes are one of the most devastating natural hazards, causing serious disasters worldwide. The threat and loss caused by earthquakes has become more severe with continued urbanization and economic growth. China has many continental earthquakes, and the associated seismic hazards are a major national problem. Thus, the investigation of the earthquake source process and earthquake prediction is important and urgent. Based on field observations, seismologists have revealed that earthquakes are essentially ruptures developing on faults, including initiation, propagation, and termination phases. However, due to the infrequency of large earthquakes, the impenetrability of the Earth, and the complexity of the source process, the earthquake governing mechanism has still not been completely resolved. This has significantly hampered earthquake prediction, the most important practical application of earthquake studies. Since the famous stick-slip experiment, laboratory investigations of earthquakes have become a popular method to uncover the mysteries of earthquakes. In the laboratory, researchers can perform highly instrumented and controlled laboratory experiments to circumvent the challenges faced in field studies. Researchers can nucleate as many controlled laboratory earthquakes as they want, facilitating the isolation and evaluation of all possible controlling factors. In addition, advanced diagnostic methods can be applied to laboratory experiments, favoring the refined observation and statistical analysis of experimental results. Based on the results of laboratory earthquakes, significant progress has been made in understanding the earthquake initiation process and precursory phenomena. It has been reported that there is a slow and smoothly accelerating earthquake nucleation process before the dynamic rupture. During this process, several abnormalities are observed and identified as indicators of laboratory earthquakes, including precursory slip, a sequence of foreshocks and a decrease in seismic velocities. Moreover, the relationships between the precursory phenomena and final events have been established preliminarily. These results are, in general, consistent with observations from natural earthquakes. Despite this progress in experimental studies, there are still several unresolved problems that inhibit the prediction of earthquakes using existing theories. Firstly, it is not clear whether the magnitude of final events is correlated to the nucleation phase. Several studies have shown that large and small earthquakes have identical onsets, suggesting that it is not possible to infer the final size of earthquakes from the nucleation phase. Secondly, how earthquakes nucleate is still under debate: The nucleation process of earthquakes may follow either preslip model or cascade model. These uncertainties in the nucleation model complicate the monitoring of earthquake precursors. Finally, it is difficult for seismologists to distinguish the precursors of earthquakes from unrelated phenomena. For example, some slow slips are reported to precede megathrust earthquakes, but many slow slips are not followed by earthquakes. It is also difficult to tell whether the largest earthquake in an ongoing sequence has already occurred or is yet to come. Apart from the earthquake initiation process, seismologists are also concerned with the dynamics of the rupture process because it is directly related to earthquake hazards. Studies of laboratory earthquakes have improved our understanding of the dynamics of earthquakes. For example, the supershear ruptures have been conclusively demonstrated experimentally and the Burridge-Andrews mechanism has been proved as the underlying mechanism for nucleation of supershear ruptures. In addition, laboratory studies have shown that the stress state and stress drop control the earthquake rupture speeds, which is consistent with the results of numerical simulations. To improve our knowledge of the earthquake source process and thus help predict earthquakes, there is an urgent need to conduct more systematic laboratory investigations of earthquakes. Combining field observations, numerical simulations, theoretical analyses, and laboratory investigations, promises to achieve major breakthroughs in understanding of earthquake source mechanisms and to further improve earthquake prediction and seismic hazard mitigation.