Feedbacks between mineral reactions and mantle convection

固体地球和许多地表特征的演变是由内部深处的运动控制的。我们的目标是通过对矿物行为的新理解来改变我们对这些运动的理解。地幔中的岩石，即地球的外半部，尽管是固体，但也能像冰川一样流动。这种流动是由密度的差异驱动的，例如密集的物质下沉。对密度的一个控制是矿物学，所以我们需要了解对矿物变化的控制。压力是关键，例如，随着压力的增加，石墨（碳的一种形式）会转变成金刚石（碳的一种密度更大的形式）。压力随着地球深度的增加而增加，就像在深海中一样。然而，在流动系统中，压力可能不仅仅与深度有关。对矿物学的另一个控制是应力（不同方向上单位面积的不同力），当地幔变形时，应力在地幔中普遍存在。这些想法可以用云的简单类比来说明。在风平浪静的日子里，云的底部在某一特定的高度上通常是不受干扰的，在这个高度上水正在凝结。在天气比较活跃的日子里，云底会受到扰动，因为风会让它们上下飘动，水也需要一段时间才能蒸发或凝结。因此，从远处观察云的底部可以告诉我们大气中正在进行的动态。同样，在地幔中，我们可以用地震波“看到”特定水平的矿物变化。在某些地方，水平会发生横向变化，有时可以用不同的化学成分来解释。我们认为，这可能在某些地方就像对云基的影响一样，在这种情况下，观测到的水平是正在进行的动力学的印记。我们的目标是了解流动地幔中的压力和应力，并预测它们对矿物学的影响。变化的矿物学会影响密度，进而影响流动模式。变化的矿物学影响流动，而流动又影响矿物学——这被称为反馈。我们将进行四项任务来理解这些反馈。在地幔条件下（25万大气压，温度高达1800摄氏度）对矿物进行的新实验测量了不断变化的矿物性质。为了了解矿物是如何变化的，我们将检查实验产品，以发现单个颗粒内的结构和化学细节。这些细节将使我们了解原子是如何运动的，这是第二项任务所需要的信息。建立数学模型来解释实验结果。要用实验中以天为单位发生的情况来预测地幔在几百万年里发生的情况，就需要用到数学方法。将这些预测纳入到整个地幔流动的数值模型中。该模型将用于预测当大型致密物体（构造板块）沉入地幔（例如在日本和南美洲）时会发生什么，并找出矿物变化的影响。它也将被用来模拟当热的低密度物质（如夏威夷和冰岛地下）上升到地表时会发生什么。对地幔矿物学的预测将用地震产生的地震波进行测试，地震波在穿过不同密度的岩石时以不同的速度传播。由于地球上剧烈的矿物变化，地震波会发生反射和折射。计算将使我们能够测试地震波如何映射步骤3中预测的特征。我们还将收集从地球各地观测到的地震波的大量数据集，以成像发生在深处的矿物变化，并根据正在进行的流动模式来解释它们。总之，我们将产生新的地幔模型，我们将使用地震波观测进行测试，并利用它们对矿物变化和地幔流动（控制地球进化的方式）如何相互反馈产生新的见解。

The evolution of the solid Earth and many surface features are controlled by movements deep within. We aim to transform our understanding of those movements through a new understanding of mineral behaviour. Rocks in the mantle, the outer half of the Earth, can flow despite being solid, in the same way that a glacier flows. This flow is driven by contrasts in density, for example dense material sinks. One control on density is mineralogy, so we need to understand the controls on mineral changes. Pressure is key, for example, graphite (a form of carbon) transforms into diamond (a denser form of carbon) with increasing pressure. Pressure increases with depth in the Earth, in the same way as it does in the deep oceans. However, in a flowing system, pressure may not relate simply to depth. Another control on mineralogy is stress (different force per unit area in different directions), which prevails in the mantle as it deforms.These ideas are illustrated by a simple analogy with clouds. On a calm day, the bases of clouds often appear undisturbed at a particular level, above which water is condensing. On a day of livelier weather, the cloud bases can be disturbed, as the wind wafts them up and down, and it takes time for water to evaporate or condense in response. Thus, looking at the bases of the clouds from a distance tells us something about the on-going dynamics in the atmosphere. Similarly, in the mantle, we have mineral changes at specific levels which we can "see" using seismic waves. In places the levels vary sideways, sometimes explained in terms of varying chemistry. We propose that this may in some places be like the effects on the cloud bases, in which case the observed levels are an imprint of the on-going dynamics.We aim to understand the pressures and stresses in a flowing mantle and predict their effects on mineralogy. The changing mineralogy will affect density, which in turn affects the flow patterns. Changing mineralogy affects flow, and flow affects mineralogy - this is called feedback. We will undertake four tasks to understand this feedback.1. New experiments on minerals at mantle conditions (250,000 atmospheres pressure, temperatures up to 1800 C) measuring evolving mineral properties. To understand how the minerals change, we will examine the experimental products to discover the details of structure and chemistry within individual grains. These details will enable us to understand how the atoms have moved around, information needed for the second task.2. Creation of mathematical models to explain the results of the experiments. The mathematics is required to use what happens in days in the experiments to predict what happens in the mantle over millions of years.3. Taking those predictions and including them in a numerical model for flow in the whole mantle. This model will be used to predict what happens when large dense objects (tectonic plates) sink into the mantle (e.g under Japan and South America) and find out what effect the mineral changes have. It will also be used to model what happens when hot less dense material (e.g. under Hawaii and Iceland) rises towards the surface.4. Predictions of mantle mineralogy will be tested using seismic waves from earthquake, which travel at varying speeds as they pass through rocks with varying densities. Seismic waves reflect and refract due to the sharp mineral changes in the Earth. Calculations will allow us to test how seismic waves can map the features predicted in step 3. We will also collect a large data set of observed earthquake waves from across the planet to image the mineral changes occurring deep within and interpret them in terms of on-going flow patterns.In summary we will produce new mantle models that we will test using seismic wave observations and use them to produce new insights into how mineral changes and mantle flow (which controls how the Earth evolves) feedback on each other.