Investigating the long-term spatial stability of LLSVPs

地球表面是高度动态的，以地震、火山和由地表构造板块运动引起的活跃断层为特征。众所周知，板块构造的驱动因素是地幔（地核和地壳之间的一层）内的对流。在过去的几十年里，我们对构造板块是如何形成、在地表移动以及沉入地球内部的理解有了很大的进步；然而，我们仍然有关于地幔对流和驱动板块构造的基本的一级问题。地震学通过使用地震引起的地震波来生成内部的3D图像，为回答这些问题提供了最好的线索。我们从地震学中发现的一个令人困惑的现象是在非洲和太平洋的最下层地幔中存在两个大型结构。这些结构直接位于地核上，并向地幔延伸数百公里。地震波在这些构造中的传播速度比平时慢，而且它们似乎有明显的边界，这表明它们可能是由与周围地幔岩石不同的东西组成的。有趣的是，这些结构位于大量异常活火山（称为热点）之下，例如夏威夷。通过重建过去的板块运动，我们发现许多灭绝的古代火山也在这些相同的地方形成。因此，如果非洲和太平洋下面的大型构造是这些热点火山的成因，那么它们一定在现在的位置存在了很长时间，也许在过去的5亿年里。这与地幔对流的计算模型相矛盾，后者预测这种结构很容易随着构造板块运动的变化而移动。PI假设这些深层地幔结构可能与地幔的其他部分具有不同的矿物粒度，这可能导致它们具有更高的粘度。研究小组将运用数值流体动力学来检验与这些结构相关的高粘度是否会导致地幔对流在地质时期的稳定，从而解释这些地理区域上火山的长期形成。更重要的是，研究人员将研究这些非洲和太平洋构造的动力学如何指导和控制地幔对流的长期模式。这项研究的结果将为理解地幔对流如何引起和控制板块构造提供关键的见解。地震层析成像显示，在最下方的地幔（非洲和太平洋下方）存在大面积的剪切波速低于平均水平的区域，通常被称为大低剪切速度省（llsvp）。据推测，它们是由大规模的成分不均匀性（如热化学堆）引起的。由于llsvp对理解大尺度地幔对流、热输运以及地球的热化学演化具有重要作用，近年来，llsvp的性质及其动力学意义的发现引起了许多活跃的研究。有趣的是，最近的古地磁研究暗示，llsvp可能在相同的位置保持了数亿年。这是令人困惑的，因为动力学计算表明，它们应该很容易被改变的俯冲模式扫过。这是一个需要解决的至关重要的问题，在这里，我们探讨了以下问题：热化学地幔对流的合理动力学/流变概念模型能否解释llsvp如何保持空间固定（或非常缓慢移动）？大多数地球动力学计算采用温度相关的流变学，导致llsvp较弱（由于其高温），因此更容易被下涌板横向推动。然而，扩散蠕变流变学对矿物粒度高度敏感（以幂律方式），因此粒度的微小变化会导致粘度的大变化。如果llsvp在地质时间尺度上与背景地幔在成分上保持不同，就没有理由期望它们与周围地幔具有相同的平均粒度。如果LLSVPs的平均粒度仅略大于背景地幔，则LLSVPs的组成粘度会增加。初步研究表明，当与粘度的温度依赖性相结合时，会导致llsvp周围有一个强大的外壳或包膜，产生不寻常的热化学结构，其动态行为与传统的被动热化学桩完全不同。这种类型的热化学对流还没有被探索过，提出的工作将测试这样一个假设，即粒度诱导的粘度成分依赖可以导致LLSVP结构，这些结构不太容易被俯冲板块推动，因此在地质时间尺度上变得更加空间稳定。将进行广泛的地球动力学研究（通过数值模拟），以探索与传统模型相比，流变学的粒度依赖性如何影响系统的流体动力学特性。此外，还将在一个模型子集中采用粒度演化，以检验在相关地质时间尺度上，llsvp是否具有比背景地幔更大的粒度。总之，数值实验将被开发和执行，以确定粘度的粒度依赖性如何影响热化学桩的空间稳定性，以及这是否是一个动态可行的机制，以产生长寿命，稳定的llsvp位置。

The Earth's surface is highly dynamic, characterized by earthquakes, volcanoes, and active faulting caused by the movement of tectonic plates at the surface. It is well-understood that the driver of plate tectonics is convection within Earth's mantle, the layer between the core and the crust. Over the past several decades, there has been much advancement toward increasing our understanding of how tectonic plates form, move about the surface, and sink back into the Earth's interior; however, we still have fundamental, first-order questions regarding how the mantle convects and drives plate tectonics. Seismology provides the best clues to answer these questions, by using seismic waves caused by earthquakes to generate a 3D image of the interior. One puzzling observation that we find from seismology is the presence of 2 large structures in the lowermost mantle, beneath Africa and the Pacific. These structures rest directly on the Earth's core, and extend several hundred kilometers into the mantle. Seismic waves travel slower than usual through these structures, and they seem to have sharp boundaries, indicating that they may be composed of something somewhat different than the surrounding mantle rock. Interestingly, these structures underlie a large number of anomalous active volcanoes (called hotspots), such as Hawaii. By reconstructing plate motions back through time, we find that many extinct, ancient volcanoes were also formed above these same locations. Therefore, if the large structures beneath Africa and the Pacific are the cause of these hotspot volcanoes, they must have been in their present location for a long time, perhaps for the past 500 million years. This is contradictory to computational models of mantle convection that predict that such structures would be easily moved by changing tectonic plate motions. The PI hypothesizes that these deep mantle structures may have a different mineralogical grain size than the rest of the mantle, which could cause them to have a higher viscosity. The team will perform numerical fluid dynamics to examine whether the higher viscosity associated with these structures could lead to a stabilizing of mantle convection currents over geologic time to explain the long-lived creation of volcanoes over these geographic areas. More importantly, the investigators will examine how the dynamics of these African and Pacific structures can guide and control the long term patterns of mantle convection. Results from this study will provide critical insight into understanding how mantle convection causes and controls plate tectonics.Seismic tomography reveals the presence of large regions in the lowermost mantle (beneath Africa and the Pacific) that exhibit lower-than-average shear wave velocity, commonly referred to as the Large Low Shear Velocity Provinces (LLSVPs). It has been hypothesized that they are caused by large-scale compositional heterogeneity (e.g., thermochemical piles). Discovering the properties and dynamical implications of the LLSVPs has involved much active research in recent years due to their critical role toward understanding large-scale mantle convection, heat transport, and thermal and chemical evolution of the Earth. Interestingly and somewhat paradoxically, recent paleomagnetic research has hinted that LLSVPs may have remained in the same positions for hundreds of millions of years. This is perplexing because dynamical calculations indicate that they should be easily swept around by changing subduction patterns. This is a critically important issue to address, and here, the following question is explored: Can a reasonable dynamical/rheological conceptual model of thermochemical mantle convection explain how LLSVPs could remain spatially fixed (or very slow moving)? Most geodynamical calculations employ a temperature-dependent rheology, resulting in LLSVPs being weaker (due to their high temperature) and therefore, easier to be laterally pushed around by downwelling slabs. However, diffusion creep rheology is highly sensitive to mineral grain-size (in a power-law manner), so small changes in grain-size can lead to large changes in viscosity. If LLSVPs have remained compositionally distinct from the background mantle over geologic timescales, there is no reason to expect that they would have the same average grain-size as the surrounding mantle. If the average grain-size of LLSVPs is only slightly larger than that of the background mantle, LLSVPs would have an increased compositional viscosity. Preliminary work has shown that when combined with temperature-dependence of viscosity, this leads to a strong rind or envelope surrounding the LLSVPs, producing unusual thermochemical structures that dynamically behave quite differently than conventional, passive thermochemical piles. This type of thermochemical convection has not been explored, and the proposed work will test the hypothesis that grain-size induced compositional dependence of viscosity can lead to LLSVP structures that are less easily pushed around by subducting slabs and therefore, become more spatially-stable over geologic timescales. An extensive geodynamical study (through numerical modeling) will be performed to explore how grain-size dependence of rheology influences the fluid dynamical properties of the system, compared to traditional models. Additionally, grain-size evolution will be employed in a subset of models, to examine whether it is feasible for LLSVPs to have larger grain size than background mantle over the relevant geological timescales. In summary, numerical experiments will be developed and performed to determine how grain-size dependence of viscosity influences the spatial stability of thermochemical piles and whether this is a dynamically feasible mechanism to produce long-lived, stable positions of LLSVPs.