The Earths's Core: Dynamics and Reversals

了解地球磁场产生的物理机制是当今突出的科学挑战之一。古地磁记录提供了地球磁场的长期历史，揭示了磁场具有一定极性的漫长时期，其间穿插着磁场反转的相对较短的时期。理解为什么地球磁场表现出这种特征行为，只能来自于对维持磁场不受其衰减趋势影响的过程的充分理解——地球动力机制。地壳之下的地球内部有三个不同的区域：一个以铁为主的固体内核；液态金属外芯；以及导电的地幔，其中的运动只能在极长的时间尺度上发生。因此，发电机位于外核，是由该区域的运动通过感应维持磁场产生的。对于外核的运动，最被广泛接受的理论是，它们是热对流和成分对流共同作用的结果。虽然控制地核动力学的方程是已知的，但由于所涉及的无量纲参数的极值，它们不能轻易求解。然而，有了今天极其强大的并行处理器计算机，有可能在某种程度上接近真正的参数体系，而且，至关重要的是，然后获得对所涉及的物理学的新见解，并随后导致新的物理解释。因此，我们建议通过大规模并行计算机上的数值模拟来研究旋转热对流驱动的发电机作用。先前对这个问题的研究表明，在某些参数条件下，磁场是小规模的，因此不会让人想起地球的偶极场，而如果旋转速度足够快，那么对流就会组织成相干列，这些列可以产生强大的大规模磁场。据推测，像地球这样的发电机，在很长一段时间内保持一个极性，但也经历间歇性的反转，位于这些小型和大型发电机之间的边界上。目前，人们对这两种发电机之间的转换性质知之甚少。我们的第一个目标是理解平面层几何中的这种转换，这在计算上是非常有效的，并且将允许对控制问题的三维参数空间进行彻底的探索。然后，利用平面层问题提供的知识，我们将在更现实，但计算要求更高的球壳几何中进行一系列重点计算。任何发电机计算的一个关键方面都涉及所施加的边界条件的性质——温度、速度和磁场。在地球本身，这些是复杂的，因此，了解各种条件的含义是非常重要的。例如，缓慢变化的热通量是否会影响发电机机制的性质，甚至可能影响逆转的模式？最后，由于现在有相当大的计算能力，我们希望能够进行足够长的运行，以便产生倒转的统计数据，从而能够与地球磁场的真实统计数据进行直接比较。

Obtaining an understanding of the physical mechanisms responsible for the generation of the Earth's magnetic field is one of today's outstanding scientific challenges. Paleomagnetic records provide a long history of the Earth's field, revealing long epochs in which the magnetic field is of a certain polarity, interspersed with relatively short periods during which the field reverses.Understanding why the Earth's magnetic field exhibits this characteristic behaviour can only come from a full understanding of the processes that maintain the magnetic field against its tendency otherwise to decay -- the geodynamo mechanism. The interior of the Earth, beneath the crust, has three distinct regions: a solid, predominantly iron, inner core; a liquid metal outer core; and an electrically conducting mantle, in which motions can occur only over extremely long time scales. The dynamo is thus located in the outer core, and results from the motions in this region maintaining the magnetic field via induction. The most widely accepted theory for the motions of the outer core is that they result from a combination of thermal and compositional convection.Although the equations governing the dynamics of the Earth's core are known, they cannot be readily solved, owing to the extreme values of the dimensionless parameters involved. However, with today's extremely powerful, parallel processor computers, it is possible to go some way towards the true parameter regime and, crucially, then to obtain new insights into the physics involved, and, subsequently, to lead to new physical explanations.We therefore propose to investigate, via numerical simulations on massively parallel computers, dynamo action driven by rotating thermal convection. Previous studies of this problem have revealed that in certain parameter regimes the magnetic field is small-scale, and hence not reminiscent of the Earth's dipolar field, whereas if the rotation rate is sufficiently rapid then the convection is organised into coherent columns, and these can generate a strong large-scale magnetic field. It has been conjectured that dynamos such as the Earth's, that maintain one polarity for a long period but also undergo intermittent reversals, lie on the boundary between these small- and large-scale dynamos. Currently little is known about the nature of the transition between these two types of dynamo. Our first aim is to understand this transition in a plane-layer geometry, which is computationally very efficient and will allow a thorough exploration of the three-dimensional parameter space governing the problem. Then, with the knowledge afforded by the plane layer problem, we shall conduct a series of focused computations in the more realistic, but computationally more demanding, spherical shell geometry.One of the crucial aspects of any dynamo calculation concerns the nature of the imposed boundary conditions -- on the temperature, the velocity and the magnetic field. In the Earth itself these are complex, and it is therefore very important to understand the implications of the various conditions. For example, will a slowly changing heat flux affect the nature of the dynamo mechanism and maybe the pattern of reversals?Finally, with the considerable computational power now available, we hope to be able to perform sufficiently long runs so as to produce statistics of reversals, thus allowing a direct comparison with the true statistics of the Earth's magnetic field.