Geodynamically relevant anisotropic tomography

该奖项是根据2009年美国复苏和再投资法案（公法111-5）资助的。地震层析成像是成像地球内部结构的最佳工具之一，这是其他方法无法实现的。将这些模型扩展到包括地震速度的各向异性（即方向依赖性），可以更直接地描绘地球深部的动态过程。虽然已经开发了许多层析模型，包括地震各向异性，但很少有反演各向异性的强度和方向，以便与地球动力学模型进行直接比较，后者根据地幔粘度和其他流变特性的假设来预测地幔对流的模式。这项研究的长期目标是开发全球以及更详细的地震各向异性区域模型，既可以阐明地幔的动态过程，也可以与地球动力学预测模型进行直接比较。该应用程序的目标是测试和应用一种新的方法，利用表面波数据的三维有限频率建模来模拟各向异性地震结构的强度和最佳拟合方向，这是实现这一长期目标的下一步。我们期望，当这种方法与剪切波分裂强度数据的有限频率建模相结合时，将产生允许与地球动力学流动模型直接比较的模型，以及允许相对稀疏的区域数据的各向异性建模。我们计划通过实现以下具体目标来证明这种方法的潜力。(1)验证面波模拟方法在利用已知各向异性模型计算的大型合成数据集上能够准确再现已知模型。(2)将剪切劈裂强度测量纳入模型，期望在空间灵敏度和对不同模型参数的相对灵敏度上实现互补。(3)将该方法应用于智利海脊俯冲项目数据集。这种临时部署的宽带地震仪器在2004年至2007年间对一个构造感兴趣的区域进行了采样，对这个相对稀疏的采样区域进行建模，既提出了挑战，也提供了机会，证明了所提出方法的优势。所提出的工作是创新的，因为它利用了PI刚刚开发的表面波数据建模的新方法。该方法有望使各向异性强度和方向的直接建模超越传统的各向异性建模。这种方法非常适合与地球动力学预测进行直接比较，从而提供了一种独特的方法来成像地幔的流动模式。这项工作将有助于促进发现和理解，同时通过培训至少一名博士生，以及潜在的一名或多名硕士和本科生，促进教学，培训和学习，并将这些项目纳入课堂作业。它还将通过优化一个新的高性能计算集群，为其他研究小组提供用于研究和教育的基础设施。最后，我们期望通过在国际会议和同行评议期刊上的介绍，广泛传播结果。在这项研究中开发的计算机代码将通过PI的网站提供。

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5). Seismic tomography is one of the best tools for imaging the interior structure of the Earth, which is inaccessible by other means. Expanding these models to include anisotropy (i.e. directional dependence) of seismic velocities can more directly image the dynamic processes of the deep Earth. While many tomographic models including seismic anisotropy have been developed, there are few that invert for both strength and orientation of anisotropy in order to allow for direct comparison with geodynamic models, which make predictions about the patterns of mantle convection based upon assumptions of the mantle's viscosity and other rheological properties. The long term goal of this research is to develop global as well as more detailed regional models of seismic anisotropy that can both illuminate the dynamic processes of the mantle, as well as allowing for direct comparison with geodynamically predicted models. The objective of this application, which is the next step towards this long-term goal, is to test and apply a new approach of utilizing 3D finite-frequency modeling of surface wave data to model both the strength and best-fitting orientation of anisotropic seismic structure. We expect that this approach, when combined with finite-frequency modeling of shear wave splitting intensity data, will produce models which allow for direct comparison with geodynamical flow models, as well as allowing for anisotropic modeling of relatively sparse regional data. We plan to demonstrate the potential of this approach by pursuing the following specific aims. (1) Verify that the surface wave modeling method can accurately reproduce known models when applied to a large synthetic dataset calculated with known models of anisotropy. (2) Incorporate shear splitting intensity measurements into the model, expected to be complementary in terms of spatial sensitivity and relative sensitivity to different model parameters. (3) Apply the approach to a dataset from the Chile Ridge Subduction Project. This temporary deployment of broadband seismic instruments sampled a region of tectonic interest between 2004 and 2007, and the modeling of this relatively sparsely sampled region presents both a challenge and an opportunity to demonstrate the strengths of the proposed approach. The proposed work is innovative because it takes advantage of a new method of modeling surface wave data that has just been developed by the PI. This method is expected to allow for direct modeling of the strength and orientation of anisotropy beyond what has been traditionally obtained in anisotropic modeling. This approach is ideally suited for direct comparison with geodynamic predictions, allowing for a unique method of imaging the flow patterns of the Earth's mantle.This work will contribute to advancement of discovery and understanding while promoting teaching, training and learning through the training of at least one Ph.D. student, as well as potentially one or more Masters and undergraduate students, and the incorporation of such projects into class work. It will also supply for the enhancement of infrastructure for research and education, through the optimization of a new HPC cluster that will be used by other research groups. Finally, we expect the broad dissemination of results through presentation at international meetings and peer-reviewed journals. Computer codes developed in this research will be made available through the PI's website.