CAREER: High-resolution Simulations of Subduction Along the Pacific Rim of Fire

板块构造理论预测地球的外层，即岩石圈，是由刚性构造板块组成的（大约100公里厚）。这些板块彼此相对运动，以1-20厘米/年的速度运动。构造板块的相互作用以会聚、发散和剪切运动为特征，大部分变形（地震、火山活动和造山活动）集中在板块边界。俯冲带是在密度较大的海洋板块向下弯曲，离开地球表面，并下降到下面的地幔的汇合板块边界。地球上大部分的俯冲带都发生在环太平洋火山带，其特点是在板块相互滑动的地方发生大规模地震活动，在板块上部由于俯冲板块进入地幔时释放的流体而形成弧状火山。该提案利用数据驱动模型设计、高性能计算和三维（3D）虚拟现实来构建环太平洋火山带的高分辨率3D模型。环太平洋火山带的地球动力学模型将研究出现在俯冲带边缘的一类新火山，而不是位于俯冲板块上方的典型弧型火山。此外，地球动力学模型将研究板块驱动的地幔流，解决构造板块和地幔之间耦合的突出问题。在更广泛的影响方面，PI将与位于伊利诺伊州芝加哥阿德勒天文馆的空间可视化实验室合作，开发环太平洋火山带俯冲的3D可视化。这些可视化将使数据探索成为可能，因为人们可以通过一系列穿越地球的虚拟航行来游览板块构造边界。此外，太平洋两侧的俯冲带形成了影响20多个国家的主要构造危险，包括主要的人口中心，使得该地区的高分辨率模拟可能会产生巨大的社会影响。一名研究生和一名本科生将被训练并积极参与这项研究。汇聚板块边缘的俯冲，以下沉的海洋岩石圈和亚平行的海沟和弧状火山为特征，通常被提炼成二维（2D）范式。然而，现代俯冲系统包含不连续的板块，这些板块不是无限长的，超过一半的板块与另一个板块相交或直接相互作用，使二维俯冲近似失效。此外，尽管在大多数俯冲带中，弧状火山确实在深处跟踪俯冲板块，但异常火山也有系统地发生，不是在板块上方，而是在板块边缘的远端，形成了一种无法用二维俯冲范式解释的模式。此外，岩石变形实验表明，大部分上地幔是由非线性流变控制的，这表明在相交的板块上，板块驱动的流动不是二维牛顿框架变化的线性组合。拟议的研究将建立整个环太平洋火山带的全面三维地球动力学模型，这将允许对自然俯冲带进行系统的比较分析，并将该领域转移到三维俯冲框架中，而不是统治更广泛的研究界和本科教科书的二维范式。具体来说，数值模拟将检验两个假设。第一种假设认为，板块边缘驱动的地幔上涌是一种普遍现象，在几何上受到三维俯冲几何的约束，在物理上受到软流圈中粘性流动的质量和动量守恒对密度异常运动的约束。第二种假设是橄榄石位错蠕变变形机制的剪切减薄效应，通过允许俯冲带附近的局部上地幔与更大范围的地幔环流模式分离，限制了观测到的俯冲带内非板块运动定向剪切波分裂的横向程度。数据驱动的环太平洋火山带三维模型将为约束这些现象提供理想的系统，因为地理参考模型可以在空间上将软流圈的上升流与环太平洋火山带异常火山的特定观测位置联系起来，并预测非线性地幔流动的程度，这可以直接与该地区的横波分裂观测结果进行比较。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The theory of plate tectonics predicts the outer layer of the Earth, the lithosphere, is composed of rigid tectonic plates (approximately 100 km thick). These plates are in motion with respect to one another, moving at speeds on the order of 1-20 cm/yr. The interaction of the tectonic plates is characterized by convergent, divergent, and shearing motion, with the majority of deformation (earthquakes, volcanism, and mountain building) concentrated at the plate boundaries. Subduction zones are convergent plate boundaries where the denser oceanic plates bend downward, leaving the Earth’s surface, and descend into the underlying mantle. The majority of the Earth’s subduction zones occur along the Pacific Rim of Fire and are characterized by large-scale earthquake activity where plates slide past one another and by arc volcanoes in the upper plate due to fluids released by the subducting slab as it descends into the mantle. This proposal leverages data-driven model design, high performance computing, and three-dimensional (3D) virtual reality to construct high-resolution 3D models of the Pacific Rim of Fire. The geodynamic models of the Pacific Rim of Fire will investigate a new class of volcanoes that occur at the edges of subduction zones, rather than typical arc volcanoes located above the subducted plate. In addition, the geodynamic models will examine slab-driven mantle flow, addressing outstanding questions of coupling between the tectonic plates and the mantle. In terms of broader impacts, the PI will collaborate with the Space Visualization Lab at the Adler Planetarium in Chicago, IL to develop 3D visualizations of subduction along the Pacific Rim of Fire. These visualizations will enable data exploration as people can tour the plate tectonic boundaries through a series of virtual voyages through the Earth. Furthermore, the subduction zones that flank the Pacific Ocean form major tectonic hazards affecting over 20 countries, including major populations centers, making the high-resolution simulations of this region likely to have a large societal impact. A graduate student and undergraduate students will be trained and actively involved in this research.Subduction at convergent plate margins, characterized by descending oceanic lithosphere and subparallel tracts of oceanic trenches and arc volcanoes, has commonly been distilled into a two-dimensional (2D) paradigm. However, modern subduction systems contain discontinuous slabs that are not infinitely long and over half of the slabs intersect or interact directly with another slab, invalidating the 2D subduction approximation. Furthermore, although arc volcanoes do track subducted slabs at depth in most subduction zones, anomalous volcanoes also systematically occur, not above a slab, but distal to the slab edge forming a pattern not explained by the 2D subduction paradigm. In addition, rock deformation experiments indicate that much of the upper mantle is governed by nonlinear rheology, indicating slab-driven flow at intersecting slabs is not a linear combination of variations on the 2D Newtonian framework. The proposed research will build comprehensive 3D geodynamic models of the entire Pacific Rim of Fire that will allow for systematic comparative analysis of natural subduction zones and move the field into a 3D framework of subduction, rather than the 2D paradigm that has governed the broader research community and undergraduate textbooks. Specifically, the numerical simulations will test two hypotheses. The first hypothesis is that slab edge driven mantle upwelling is a common phenomenon, constrained geometrically by the three-dimensional subduction geometry and physically by the constraints on the motion of density anomalies due to conservation of mass and momentum of viscous flow in the asthenosphere. The second hypothesis is that the shear thinning effects of the dislocation creep deformation mechanism of olivine constrains the lateral extent of non-plate motion aligned shear wave splitting in observed subduction zones, by allowing the localized upper mantle near the subduction zone to decouple from the larger-scale mantle circulation patterns. A data-driven 3D model of the Pacific Rim of Fire will provide the ideal system to constrain these phenomena, because a geographically referenced model can spatially link upwellings in the asthenosphere to specific observed locations of anomalous volcanics along the Pacific Rim of Fire, as well as predict the extent of nonlinear mantle flow which can be directly compared to shear wave splitting observations from the region.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.