Collaborative Research: Improving lower mantle seismic sampling and model resolution using multi-bounce and diffracted waves

地震产生的地震波会穿过地球的整个内部。这些波被用来成像行星内部。地震层析成像工具经常被用来产生地幔内地震波速度变化的图像（地幔大致占据了地球的外半部）。我们提出的工作旨在通过使用更多类型的地震波来改进过去的层析成像研究，这些地震波将以一种新的和新颖的方式进行测量，以提高准确性。其中一些波包括地震S波，它从地核反弹并多次返回地表。这使我们能够对地球的地震波进行采样？在断层扫描研究中，南半球的采样比北半球更差。提高地幔结构的分辨率对于提高我们对全球内部过程的本质的理解是重要的，包括地幔的对流动力学和地球的演化。同样令人感兴趣的是，在地幔底部的两个巨大的斑点状结构内提高层析成像的清晰度，这两个结构的大小相当于大陆，向上延伸到地幔中至少1000公里（科学上称为大低剪切速度省，或llsvp，它强调通过这些结构的剪切波的速度大大降低）。通过地震层析成像方法对llsvp的大尺度特征进行表征，不同的研究小组得出了类似的结果。然而，模型之间的小尺度结构不同。这个为期两年的项目旨在为对较小尺度结构敏感的过程增加更多信息，即对多次反弹S和sc波以及衍射S和P波的仔细旅行时间测量。我们将在正演和逆演建模方法中使用这些方法。全层析反演将针对P和S结构。正演方法使用新数据迭代更新现有的断层扫描模型，并更好地保留更清晰的结构。在更大的地震中，可以清楚地观察到波浪反弹多达5次，即6段旅程，例如S6和ScS6（即6条S和sc路径）。多次弹跳数据非常适合改进地幔成像，因为它们允许小弧和大弧传播路径（即，地震和台站之间的大圆路径，以及沿相反方向的大圆路径，分别绕地球很长一段路）。我们将为这些长路径数据开发有限频率核。正演和反演方法的最终三维模型将用于计算三维合成地震图，以与实际数据进行比较。这将既评估模型鲁棒性，也比较正向和逆方法的解决方案。地幔的非均质性取决于温度、矿物学、相位和状态，虽然层析成像只能为我们提供当今的快照，但它可以与地球所采取的进化途径有关。近年来，人们越来越关注热点火山活动位置与大火成岩省起源位置等地表观测数据与深部构造（如llsvp）之间的关系。因此，提高地幔非均质性地震图像的分辨率使我们更接近于理解地球的结构、动力学和演化。

Earthquakes generate seismic waves that travel through the entire interior of the planet. These waves are used to image the planetary interior. The tool of seismic tomography is frequently used to produce images of the variation of seismic wave speeds within the Earth's mantle (the shell roughly occupying the outer half of the planet). Our proposed work aims to improve upon past tomographic studies by using a larger suite of types of seismic waves that will be measured in a new and novel way in order to improve accuracy. Some of the waves include seismic S waves that bounce off of the Earth's core and back to the surface, multiple times. This enables us to add seismic wave sampling of Earth?s southern hemisphere, which is more poorly sampled than the northern hemisphere in tomography studies. Improving the resolution of Earth's mantle structure is important for improving our understanding of the nature of global internal processes, including the convective dynamics of the mantle and evolution of the planet. Also of interest is improving the clarity of tomographic images within two massive blob-like structures at the base of Earth's mantle, which are continental in size and extend at least 1000 km up into the mantle (scientifically referred to as Large Low Shear Velocity Provinces, or LLSVPs, which emphasizes the large reduction in the speeds of shear waves through these structures).The characterization of the large-scale aspects of LLSVPs by the method of seismic tomography yields similar results from different research groups. However, the smaller scale structure differs between models. This 2-year project aims to add more information to the process that is sensitive to the smaller scale structure, namely careful travel time measurements of multi-bounce S and ScS waves, and also diffracted S and P waves. We will use these in both forward and inverse modeling approaches. The full tomographic inversion will be for both P and S structure. The forward approach iteratively updates existing tomography models using new data, and better preserves sharper structures. Waves bouncing up to 5 times, thus 6 legs of the journey, e.g., S6 and ScS6 (i.e., six S and ScS paths), are clearly observed for larger earthquakes. The multi-bounce data are well suited for improving mantle imaging, since they allow for both minor and major arc travel paths (i.e., the great circle path between earthquake and station, as well as along the great circle path in the opposite direction, the long way around the planet, respectively). We will develop finite frequency kernels for these long path data. Final 3D models from both the forward and inverse approaches will be used to compute 3D synthetic seismograms to compare to actual data. This will both assess model robustness as well as compare solutions for the forward and inverse methods. Mantle heterogeneity depends upon temperature, mineralogy, phase, and state, and while tomographic imaging only provides us with a present-day snapshot in time, it can be related to the evolutionary pathway Earth has taken. Recently, there has been increased attention to the relationship between surface observables, such as the locations of hot spot volcanism and the origination locations of large igneous provinces, with deep structures (e.g., LLSVPs). Thus improving resolution in seismic images of mantle heterogeneity brings us closer to understanding the structure, dynamics, and evolution of our planet.