Geodynamic modelling and tomographic visibility of a mantle plume under La Réunion

德国-法国实验Rhum-rum的目标是在地震学上成像留尼汪岛热点下的地幔热柱--或它的缺失--并了解地球地幔所有深度水平上的对流结果。大部分海底和岛屿数据将在2013年底提供。我们在这里提出的工作包括两种相互关联的先验建模研究，旨在伴随着Rhum-rum的数据分析组件。在工作包1(一名博士生)中，我们提议开发一个新的留尼汪岛地幔热柱与大尺度流动相互作用的地球动力学模型，通常认为这主要是由俯冲板块驱动的，但也可能是由非洲和印度洋下活跃的上升流驱动的。在工作包2(一名博士生)中，我们将研究这种先验合理的羽流结构的层析可观测性。我们提出了一个完整的参数和不确定性研究先验(在每个深度下，每个流变学/物理参数的阈值应该仍然是可检测的？)，而不是通常后验地显示的非常局部的分辨率测试，以证明从真实数据获得的首选模型的合理性。因此，我们的地球动力学模型及其预期的层析特征可以被视为真实的预测。随着同事们从Rhum-rum数据分析中获得观测结果，这些模型将与它们进行比较，地球动力学参数将被缩小，并将运行修改后的模型。这两个工作包都包含了计算方法开发的强大组成部分，就地幔对流模拟和宽带波形反演而言。我们的首要目标是：1.总体上更好的地球动力学模型，从大的低剪切波速度省(LLSVP)升起的假设的经典羽流过程--一种目前被认为适用于大多数海洋热点的生成机制。2.清楚和全面地了解哪些结构，如果存在，可以或不能通过实际的Rhum-rum数据来解决。实际观察到的结构的直接地球动力学背景，一旦它们可用，即在项目经历其最大势头和科学界最大兴趣的时刻。

The goal of the German-French experiment RHUM-RUM is to seismologically image a mantle plume - or its absence - under the hotspot of La Réunion, and to understand the results in terms of convective flow at all depth levels of the Earth mantle. Most of the ocean-bottom and island data will become available in late 2013.The work we propose here consists of two kinds of interrelated, a priori modeling studies, meant to accompany the data analysis components of RHUM-RUM. In Work Package 1 (one Ph.D. student) we propose to develop a new geodynamic model of the Réunion mantle plume interacting with large-scale flow, which is commonly considered to be largely driven by subducted slabs, but possibly also by active upwellings under Africa and the Indian Ocean. In Work Package 2 (one Ph.D. student), we will investigate the tomographic observability of such a priori reasonable plume structures.We propose a full-fledged parameter and uncertainty study a priori (Down to which threshold of each rheological / physical parameter should a plume still be detectable, at each depth?), rather than the very topical resolution tests that are typically shown a posteriori in order to justify a preferred model obtained from the real data. Our geodynamic models, and their expected tomographic signatures, can therefore be regarded as a true predictions. As observations become available from the RHUM-RUM data analyses by colleagues, the models will be compared to them, geodynamical parameters will be narrowed down, and modified models will be run. Both work packages contain strong components of computational method development, in terms of mantle convection simulations and broadband waveform inversion.Our overarching objectives are:1. Better geodynamical models in general, of the process of a hypothesized classical plume rising from a Large Low Shear wave Velocity Province (LLSVP) - a generation mechanism currently thought to apply to most oceanic hotspots. 2. A clear and comprehensive understanding of which structures, if present, could or could not be resolved by the actual RHUM-RUM data.3. An immediate geodynamical context for actually observed structures, once they become available, i.e., at a moment when the project experiences its maximum momentum and greatest interest from the scientific community.