Rheological control of the dynamics of Earth's transition zone

地球表面的山脉和海洋盆地是内部对流的结果。在地核和地幔中产生和储存的热量，驱动着一个巨大的对流引擎，使物质在整个内部和表面移动。地球上的地理、地质甚至生命都是这一过程的结果。在最大的尺度上，我们可以确定地球上两个主要的边界，作为对流的障碍。很少有物质穿过岩石地壳和地幔之间的边界，以及流动的海洋和大气。在地表下2891公里的流体外核和固体地幔之间也有类似的物质交换。热量通过传导而不是对流穿过这两个边界。地幔中的第三个屏障更加神秘。在410 ~ 660 km深度的过渡带中，地幔矿物的一系列结构重排导致地幔性质的巨大变化。在这一边界之下，下地幔似乎会变形，从而对流，比上地幔困难得多。有人认为这是地球对流的另一个障碍。然而，最新的证据表明，大量的物质渗透到过渡区。该项目旨在探索对流如何在地球上长时间运行，并对过渡区的存在作出反应，并回答以下问题：有多少物质越过边界，对流的部分障碍对地球演变的影响是什么？该项目的一个关键挑战是准确描述地幔变形的过程。正是这些变形过程使得地幔对流研究起来很有趣，建模起来也很有挑战性：地幔变形是多尺度的（原子之间、晶体缺陷之间和相邻矿物颗粒之间的相互作用都很重要）和多物理学（涉及的过程最好由不同的理论描述，如流动流体和电子量子力学）。地幔对流的第二个迷人的方面是在地表环境和地幔动力学之间作用的许多反馈过程。例如，从海洋中向地幔中加入水会导致剧烈的削弱。在大规模的模拟中，地幔通常被视为在百万年的时间尺度上缓慢流动的非常粘稠的糖浆。然而，地幔实际上并不是这样工作的：它是一块岩石。事实上，地幔是通过其晶体结构中原子缺陷的运动而“流动”的，这个过程可以用原子尺度的计算机模拟来描述。岩石和糖浆之间的区别是最重要的，因为岩石有能力存储其先前变形的历史。当岩石变形时，它们的性质也会发生变化，比如说，改变其晶体中的颗粒大小或原子缺陷的浓度。与糖浆不同，这意味着地幔现在的变形取决于它过去的变形。该项目将首次将这些过程直接纳入地幔对流的模拟中。如果没有这种能力，人们就不得不严重质疑我们对地球热机如何运作的理解，事实上，我们从对流模拟中做出的推论忽略了岩石行为的一个关键方面。

Mountain belts and ocean basins on the Earth's surface are a consequence of convection of the interior. Heat, generated and stored in the core and mantle, drives a gigantic convective engine which moves material throughout the interior and across the surface. Geography, geology and even life on Earth are the outcome of this process. On the largest scale we can identify two major boundaries in the Earth that act as barriers to convection. Very little material moves across the boundary between rocky crust and mantle, and the fluid oceans and atmosphere. There is a similarly trivial exchange of matter between the fluid outer core and solid mantle 2891 km beneath the surface. Heat crosses these two boundaries by conduction, not convection. A third barrier within the mantle is more enigmatic. In the transition zone at depths between 410 and 660 km a series of structural rearrangements in mantle minerals leads to dramatic changes in the properties of the mantle. Below this boundary the lower mantle appears to deform, and thus convect, with much greater difficulty than the upper mantle above it. Some have argued that this acts as another barrier to convection in the Earth. However, the latest evidence shows that a significant quantity of material penetrates the transition zone. This project is designed to explore how convection operates in the Earth on a long time-scale and in response to the presence of the transition zone and answer questions like: How much material crosses the boundary and what are the consequences of a partial barrier to convection for Earth's evolution?A key challenge in this project is to accurately describe the process that allows the mantle to deform. It is these deformation processes that make mantle convection interesting to study and challenging to model: mantle deformation is multi-scale (with interactions between atoms, between crystal imperfections and between adjacent mineral grains all being important) and multi-physics (involving processes best described by theories as diverse as those of flowing fluids and quantum mechanics of electrons). A second fascinating aspect of mantle convection are the many feedback processes acting between the surface environment and the dynamics of the mantle. For example, adding water from the oceans to the mantle causes dramatic weakening. In large-scale simulations the mantle is often treated like extraordinarily sticky treacle slowly flowing on a million-year timescale. However, the mantle does not actually work like this: it is a rock. The mantle in fact 'flows' by the motion of atomic imperfections in the structure of its crystals and this process can be described using atomic scale computer simulations. The difference between rock and treacle is most important because of the ability of rocks to store the history of their previous deformation. As they deform rocks change their properties by, for example, changing the grain size or the concentration of atomic imperfections in their crystals. Unlike treacle this means that how the mantle deforms now depends on how it has deformed in the past. The project will, for the first time, incorporate these processes directly into simulations of mantle convection. Without this ability, one has to seriously question our understanding of how the Earth's heat engine operates and, indeed, the inferences we make from simulations of convection that ignore a key aspect of the behaviour of rocks.