Coherence-based imaging and inversion on large-scale seismic arrays

对地球地下结构和成分的详细了解可以用于许多重要的应用，从地表附近工程场地的风险评估到地热或碳氢化合物储层的勘探，以满足能源需求，以及回答大陆甚至全球尺度上的基本地球动力学问题。利用地震的自然地震活动性或人工激发可控震源，地震方法被证明是解决所有这些问题的重要工具。然而，在全球地震地震学体系中，地震速度等地下特性的层析反演发挥了核心作用，而长期以来，工业驱动的可控震源局部地震学在油气勘探中一直以聚焦波场的后向散射（即反射或衍射部分）获得尽可能清晰的结构图像为主要目标。随着全波形反演的出现和跨越地球表面大片地区的接收网络的安装，这些领域的技术发展开始自然地相互接近。除了全波形的反演之外，目前大规模地震阵列的快速增长表明了成像方案的应用，这与油藏勘探类似，可能会导致对仍然模糊的地幔和地核结构的更丰富，互补的图像。虽然近年来使用常规地震深度成像技术获得了第一个有希望的结果，但本项目的目的是最大限度地利用地震地震学中获得的数据中日益增加的冗余。通过采用来自工业地震勘探的尖端数据驱动叠加方案，我的目标是补充和支持现有的宽带层析成像方法，并为最近在北美完成的密集采集USarray活动提供更深层的大规模结构图像，这对两个领域的未来工作流程都有有趣的影响。

The detailed knowledge of the structure and constituents of the Earth's subsurface allows for many important applications, ranging from risk assessments on engineering sites near the surface to the exploration of geothermal or hydrocarbon reservoirs to meet energy demands and answering fundamental geodynamical questions on continental or even global scales. Using either the natural seismicity of earthquakes or artificially excited controlled sources, the seismic method proved to be an essential tool for all these problems. While in the global regime of earthquake seismology, however, the tomographic inversion of subsurface properties like seismic velocities has played the central role, the industry-driven local seismology of controlled sources in hydrocarbon exploration for a long time had the main goal of a sharpest possible structural image by focusing the backscattered, i.e. reflected or diffracted portion of the wavefield. With the advent of full-waveform inversion and the installation of receiver networks that span large regions of the Earth's surface, technical developments in these fields begin to naturally approach each other. Besides the inversion of full waveforms, the rapid growth of current large-scale seismic arrays suggests applications of imaging schemes, which, similar to reservoir exploration, may lead to a richer, complementary picture of the still vaguely known structure of the Earth's mantle and core. While first promising results have been obtained in recent years using conventional seismic depth imaging techniques, the aim of this project is to make the maximum use of the increasing redundancy in the data acquired in earthquake seismology. By adapting a cutting-edge data-driven stacking scheme from industrial seismic exploration, I aim at complementing and supporting the established approach of broadband tomography with large-scale structural images of the deeper interior for the very recently completed densely acquiring USarray campaign in North America, promising interesting implications for future workflows in both fields.