Preliminary Studies On Mechanical Models Of Earthquake Precursors And Earthquake Prediction

Strong earthquakes can be catastrophic to society, and precise prediction of large earthquakes is very important for seismic hazard reduction. In this study, we try to develop methods to quantify physical parameters associated with seismogenic processes, and use that to experiment enhancing the power of short-to-intermediate-term earthquake prediction. Specific works are the following.1)Improvement on Load/Unload Response Ratio methodWe replace the circular critical region usually adopted in LURR practices with an area within which the Coulomb stress change would affect the potential seismogenic fault of a future large event the most, to enhance the sensitivity of the LURR method. The Coulomb stress change before a hypothetical earthquake is calculated based on a simple back-slip dislocation model of the event. The results given by applying the approach to the Mw > 6.5 earthquakes occurred in California over the last twenty years suggest that the predictive power of the LURR method can be enhanced significantly by combining the LURR method with our choice of identified area with increased Coulomb stress.We develop a probabilistic approach for assessing earthquake potential based on the LURR method. In the approach, the triggering probability of a future earthquake given the confidence level of an anomalous LURR is quantitatively evaluated associated with stochastic distribution of LURR. By applying the approach to 40 large earthquakes occurred in the Chinese Mainland, southern California, and the Kato region in Japan, we find a linear increase of earthquake potential along with the confidence level of precursory LURR anomaly. The research results indicate that our probabilistic approach can provide a useful tool to quantitatively evaluate the possibilities of critical triggering of the future earthquakes.