How does crust form at arcs? - a Multidisciplinary Study of the Lesser Antilles Volcanic Arc

大陆地壳的形成深刻地影响了生命的进化、大气、水圈和生物圈的化学成分以及资源的分布，因此了解它是如何形成的是一个一级地球化学和地球物理问题。目前，或许在整个地球历史的大部分时间里，大多数新地壳都是在俯冲带上方产生的，因此俯冲带岩浆活动与地壳生长密切相关。然而，在俯冲带生成的新未成熟地壳的成分主要受玄武岩岩浆流的控制，而大陆地壳的硅质和不相容的微量元素含量则高于此（~安山岩）。为了从弧壳变成真正的大陆壳，还需要在俯冲期间（如深部堆积的分异和分层）或在俯冲之后（如地壳内部混合或碰撞时深部地壳的移除）在弧壳中发生的附加过程。揭开这些潜在的重要机制的关键在于真正了解俯冲带（上板块）地壳的结构和组成。为此，我们提出了一种结合岩石学、岩石学实验、地球化学和地震学的新方法来表征大洋内岛弧（小安的列斯群岛）的地壳。将地震约束与岩石学控制相结合是一种特别创新的方法，可以通过广泛的捕虏体收集和非常大的高质量地震数据集来实现。为了清楚地解释这种整合将如何完成，我们提供以下类比：假设你有一些拼图碎片（捕虏体），并想找出整个拼图的图像（在本例中是小安的列斯弧地壳的地球化学/矿物学剖面）。拼图片是图片的代表-所有缺失的都是相同的类型，你已经有了。如果你把拼图想象成一个x-y网格，那么岩石学实验，以及来自矿物成分的地球气压测量，可以用来告诉你拼图碎片占据了几行（捕虏体来自多深）。然后，地震数据可以作为一种“模式”，告诉我们哪些部分在二维x-y网格中位于何处。首先，这提供了地震和岩石学数据之间的验证（两种方法确定的深度需要一致）。其次，既然我们现在知道了哪些地震数据对应于哪种捕虏体类型，那么我们就可以使用地震数据来填充图片的其余部分，并建立一个沿弧的地壳剖面。这些方法之间的互补性是关键；地震资料不能告诉我们确切的矿物组合，而只能说明矿物的组合。捕虏体，反过来，只能提供下地壳的例子，并且只能通过将它们的地震特性（从捕虏体的矿物学计算）与观测到的地震剖面联系起来，来确定它们作为给定地壳深度的代表。因此，通过将捕虏体数据、实验数据和地震数据结合起来，我们将能够生成一个沿整个弧形的地壳结构模型，这将为地壳是如何在俯冲带之上形成提供新的线索，这是大陆地壳最终形成的关键阶段。该项目汇集了一个多学科、多机构的国际团队。最近在布里斯托尔的采样和实验活动使这项工作特别及时，它建立在丰富的现有数据和先前的研究基础上，直接与正在进行的、得到大量资助的研究链相结合。因此，所要求的资源仅代表了大规模和长期的地球物理分析活动成本的一小部分，否则需要揭示约30x400公里弧壳剖面的结构和成分。

The creation of continental crust has profoundly influenced the evolution of life, the chemistry of the atmosphere, hydrosphere and biosphere and the distribution of resources, so knowing how it is made is a first-order geochemical and geophysical problem. At present, and probably throughout much of Earth history, most new crust is generated above subduction zones, so subduction zone magmatism and crustal growth are intimately linked. However, the composition of the new, immature crust generated at subduction zones is controlled by a largely basaltic magma flux, whereas the continental crust is more silica and incompatible trace element rich than this (~andesitic). Additional processes occurring in the arc crust during subduction (such as differentiation and delamination of deep cumulates) or subsequent to subduction (such as intra-crustal mixing or removal of deep crust during collision) are needed in order to go from arc crust to true continental crust. The key to unravelling these potentially important mechanisms lies in actually knowing the structure and composition of the (upper plate) crust at subduction zones. To that end, we propose a novel approach integrating petrology, petrological experiments, geochemistry and seismology to characterise the crust of an intra-oceanic island arc (The Lesser Antilles).The integration of seismic constraints with the petrological controls is a particularly innovative approach, made possible by the availability of an extensive xenolith collection, and a very large high-quality suite of seismic data. To explain clearly how this integration will be done, we offer the following analogy: Suppose you had a handful of jigsaw pieces (the xenoliths) and wanted to find out what picture was made by the whole jigsaw (in this case the geochemical/mineralogical profile of the Lesser Antilles Arc crust). The jigsaw pieces are representative of the picture - all the missing ones are the same types as the ones you already have. If you think of the jigsaw as an x-y grid, then the petrological experiments, along with geobarometry from mineral compositions, can be used to tell you what rows the jigsaw pieces occupy (how deep the xenoliths come from). The seismic data can then be used as a "pattern" telling us which pieces go where in a 2-dimensional x-y grid. Firstly this provides validation between the seismic and petrologic data (the depths determined by both methods need to agree). Secondly, since we now know what seismic data correspond to which xenolith type, we can then use the seismic data to fill in the rest of the picture and build an along-arc crustal section. The complementarity between these approaches is key; the seismic data cannot tell us the exact mineral assemblage, but only permitted combinations of minerals . The xenoliths, in turn, can only provide examples of the subjacent crust, and can only be confirmed as representative of a given crustal depth by tying to their seismic properties (calculated from the xenoliths' mineralogies) to the observed seismic profiles. So by drawing together the xenolith data, the experimental data and the seismic data we will be able to generate a model of the crustal structure along the entire arc, which will shed new light on just how crust is built above subduction zones, a critical stage in the ultimate generation of continental crust. The project brings together a multidisciplinary and multi-institutional international team.The recent sampling and experimental campaign at Bristol makes this effort particularly timely and it builds on a wealth of existing data and prior research, integrating directly with ongoing, substantively funded research strands The resource requested therefore represents a mere fraction of the cost of the large and protracted geophysical-analytical campaign that would otherwise be needed to reveal the structure and composition of a ~30x400km section of arc crust.