Fiber-optic volcano-tectonic fault imaging with coherent wavefields and deep learning

对人口稠密的火山地区的地下进行高分辨率地球物理成像对于评估和减轻潜在危险至关重要。最近出现的分布式声传感(DAS)允许利用光纤对连续的多通道地震应变数据进行经济高效的测量，特别适合于城市地区。在这个项目中，我提议将为多道地震数据设计的既定处理方案与相干波场增强和分离的深度学习技术结合起来，以便在(意大利)埃特纳火山东侧获取的光纤测量中识别和成像火山-构造相互作用驱动的小规模断层。

High resolution geophysical imaging of the subsurface in densely populated volcanic regions is crucial for assessing and mitigating potential hazards. The recently emerging distributed acoustic sensing (DAS) allows the cost-efficient measurement of continuous multi-channel seismic strain data with fiber-optic cables and is particularly suited for urban areas. In this project, I propose to combine established processing schemes designed for multi-channel seismic data and deep learning techniques for coherent wavefield enhancement and separation in order to identify and image small-scale faults driven by volcano-tectonic interaction in fiber-optic measurements acquired on the eastern flank of Mount Etna (Italy).