论文部分内容阅读
特征地震假设是模拟大断层上随时间变化的地震复发过程的基础。但是,特征地震假设并没有有力的观测证据。鲜有断层段具有单个已破裂事件的悠长历史记录或古地震记录,加上资料与参数的不确定性,很难获取复发的分布特征。例如,科林斯湾断层系(CGFS)的强震记载超过2 000年,虽然最近300年M≥6.0地震的目录是完整的,但其中记录的独立断层段上发生的特征地震则凤毛麟角。采用基于物理的地震模拟算法可以产出长达10万年且包含超过50万个4级以上事件的地震目录。此模拟算法的主要特征是:(1)对断层系中每个断层段都采用平均的地震滑动速率;(2)采用探索式程序实现破裂生长和停止,从而得到自组织的震级分布;(3)地震震源间具有相互作用;(4)应力再分布过程中微小地震的影响。用此算法模拟科林斯湾断层系可以给出较为真实的地震时空强特征。这些特征包括强震的长期周期性、强震与小震事件的短期丛集性以及在较高震级区间偏离古登堡—里克特分布的较真实的震级分布。
Characteristic seismic hypotheses are the basis for simulating the seismic recurrence over large faults over time. However, there is no strong observational evidence for the characteristic earthquake assumptions. It is very difficult to obtain the distribution characteristics of recurrence because of the long or historical record of a single rupture event and the records of seismotectonic records with few data and parameters. For example, the strong earthquakes of the Corinthian Fault System (CGFS) have been recorded for more than 2,000 years. Although the catalog of M ≥ 6.0 earthquakes has been completed in the recent 300 years, the number of characteristic earthquakes recorded in the independent fault segments is very small. Earthquake catalogs up to 100,000 years long and containing more than 500,000 Level 4 or above events can be produced using physics-based seismic simulation algorithms. The main features of this simulation algorithm are: (1) to adopt an average seismic slip rate for each fault section in the fault system; (2) to exploit the exploratory program to achieve rupture growth and stoppage so as to obtain the self-organized magnitude distribution; (3) ) Interaction between earthquakes; (4) the effect of minor earthquakes during stress redistribution. Using this algorithm to simulate the Collins Bay fault system can give a more realistic seismic time-space characteristics. These features include the long-term periodicity of strong earthquakes, the short-term clustering of strong earthquakes and minor earthquakes and the more realistic magnitude distribution of the Gutenberg-Richter distribution at higher magnitudes.