A probabilistic approach for earthquake potential evaluation based on the load/unload response ratio method

摘要

Previous studies indicate that the occurrence of a large earthquake might be predicted by anomalous temporal increase of the load/unload response ratio (LURR), which was often defined as the ratio of Benioff strain of small earthquakes released during loading and unloading time periods, corresponding to earth tide-induced Coulomb failure stress change on optimally oriented faults. The conventional LURR anomalous evaluation usually sets a critical LURR value, above which an earthquake may occur. In this paper a probabilistic approach for the evaluation of earthquake potential based on the LURR method is developed. In the approach, the occurrence probability of a future earthquake is quantitatively evaluated by assessing the confidence level of LURR anomaly associated with its stochastic distribution. As retrospective studies, we apply the approach to investigate the time series of LURR prior to the 50M>6.3 earthquakes that occurred in the Chinese mainland and the 21M>6.0 earthquakes in southern California over the past 30 years, and find high correlation between the confidence level of the LURR anomalies and the occurrence of the large earthquakes. We then depict all the high peaks that appeared in the LURR time series, and evaluate the earthquake occurrence rate as a function of the confidence level. The research results show considerable promise that our probabilistic approach may provide a useful tool to evaluate quantitatively the occurrence possibilities of future earthquakes.