Intermediate Depth Earthquakes: Investigations of the Predominance and Cause of Subhorizontal Faults

近100年前就发现了深层地震，但它们是如何在正常压力-温度条件下发生脆性破坏的，这仍然是一个谜。根据最近的观测，深度约75-300公里的地震经常沿着亚水平断层发生，我们提出了新的假设和试验来研究这些断层的优势和原因。当板块弯曲和俯冲时，外部隆起断层是否仍然很弱？或者，化石外隆断层之所以强大，是因为它们已经愈合了吗？弯曲板中的应力场是否导致水向下抽水而引发地震？是否存在持续的断层带？断裂传播是一个等压过程吗？不同的地震观测将支持或反驳这些假设，我们选择了两个研究区域，日本和哥伦比亚，其特征和区域数据集非常适合回答这些问题。在日本，我们正在使用宽带F-Net记录来分析~500次至~4.5级地震的指向性。在哥伦比亚，我们正在与洛斯安第斯大学的German Prieto教授合作，通过重新定位和震源特征来分析2000年考卡地震群。这些分析将促进对地球科学中一个长期未解决问题的理解。在这个过程中，研究生、本科生和早期职业科学家的参与将促进教学和培训，国际合作将加强研究和教育的基础设施。

Deep earthquakes were discovered nearly 100 years ago, but how they can occur outside the normal pressure-temperature conditions for brittle failure remains a mystery. Following up on recent observations that earthquakes from ~75-300 km depth often occur along subhorizontal faults, we have developed new hypotheses and tests to investigate the predominance and cause of these faults. Do outer rise faults remain weak as the slab bends and subducts? Alternatively, are the fossil outer rise faults strong because they have healed? Does the stress field in the bending slab result in downward pumping of water that initiates earthquakes? Do sustained fault zones exist? Is rupture propagation an isobaric process? Different seismic observations would support or refute each of these hypotheses, and we have selected two study areas, Japan and Colombia, with features and regional data sets ideal for answering these questions. In Japan, we are using the broadband F-Net records to analyze the directivity of ~500 earthquakes down to magnitude ~4.5. In Colombia, we are collaborating with Professor German Prieto of the Universidad de los Andes to analyze the Cauca cluster of ~2000 earthquakes through relocation and source characterization. These analyses will advance understanding of a long-unresolved question in the Earth sciences. In the process, the participation of a graduate student, an undergraduate student, and an early-career scientist will promote teaching and training, and the international collaboration will enhance the infrastructure for research and education.