Crystal forensics: constraining the timescales of magmatic processes leading to volcanic eruptions

火山每天都在喷发，是全世界数百万人的日常危险。今天，活火山被持续监测，寻找即将爆发的迹象。然而，为了准确预测未来的火山爆发，需要对火山内部运作的全面了解，包括熔岩在火山中移动的时间尺度。岩浆是由液态岩石和称为晶体的固体颗粒组成的熔融岩石。岩浆在地壳中的迁移和最终的喷发是复杂的，然而这一过程的记录被保存在晶体中。就像你最喜欢的犯罪剧使用法医学来解决犯罪一样，我们也可以用火山爆发的矿物成分来解决犯罪问题。晶体的生命并不简单，而是一个复杂的生命，包括生长和熔化、迁移以及在不同岩浆体中停留的时期，直到最终晶体喷发，成分冻结。这些经验中的每一个都作为化学或纹理标记保存在晶体中，从而形成高度分区的晶体。这些晶体区可以用与树木年轮相同的方式来处理，树木年轮记录了树木的生长历史。各个区域的化学成分可以用来识别形成该区域的岩浆过程，我们可以利用两个相邻区域之间化学成分的差异来确定这些岩浆过程发生的时间尺度。在形成之后，两个相邻区域之间的成分差异是尖锐的，但随着时间的推移，元素的扩散（晶体所用的小颗粒的迁移）使这种成分边界变得平滑，这可以用来计算岩浆过程的时间尺度。重要的是，不同的元素以不同的速率通过晶体结构迁移，因此可以研究在喷发前几小时到几个月的时间尺度上发生的一系列岩浆过程。然而，这些晶体中的分区发生在亚微米尺度上（小于人类头发的宽度）。因此，这项研究将首次使用新一代的高分辨率二次离子质谱仪允许测量细尺度的化学分带的斜长石晶体从圣海伦山在美国和塔拉纳基在新西兰。这将使岩浆过程和这些过程直接发生在喷发之前的时间尺度得到评估。这些时间尺度，然后可以评估对已知的时间尺度的岩浆运动之前，最近爆发的圣海伦火山允许更好的模型，为未来预测火山爆发的约束。这将为火山喷发前的火山活动提供宝贵的见解，并可能给全世界生活在火山阴影下的数百万人带来巨大利益，并有助于减轻活火山的相关危害。这项研究将在布里斯托大学地球科学系与布里斯托界面分析中心、德国明斯特大学、西澳大利亚大学、美国喀斯喀特火山观测站和俄勒冈大学以及新西兰惠灵顿的维多利亚大学联合进行。

Volcanoes erupt daily and are an everyday hazard for millions of people worldwide. Today, active volcanoes are continuous monitored looking for signs of imminent eruptions. However, in order for the accurate prediction of future eruptions to be achieved a comprehensive knowledge of the internal workings of volcanoes is required, including the timescales over which molten rock moves through the volcano. Magma is molten rock composed of liquid rock and solid particles called crystals. The migration of magma through the Earth's crust and the eventual eruption is complex, however a record of this is preserved in crystals. Just as your favourite crime drama uses forensic science to solve the crime, we can do the same with the composition of minerals from volcanic eruptions. The life of a crystal is not simple but a complex one involving periods of growth and melting, migration and periods of residence in different magma bodies until eventually the crystal is erupted and the composition frozen in. Each of these experiences are preserved in the crystal as chemical or textural markers, that results in highly zoned crystals. These crystal zones can be treated in just the same way as tree rings which record the growth history of a tree. The chemical composition of individual zones can be used to fingerprint the magmatic process that formed the zone and we can use the difference in the chemical composition between two adjacent zones to determine the timescale over which these magmatic processes occurred. Immediately after formation the compositional difference between two adjacent zones is sharp, but with time diffusion of elements (migration of small particles that the crystal is made from) smoothes this compositional boundary and this can be used to calculate the timescale of magmatic processes. Importantly, different elements migrate through the crystal structure at differ rates and therefore a whole range of magmatic processes that occur on the timescales of hours to months prior to eruption can be investigated. However, the zoning in these crystals occur on a sub-micron scale (smaller than the width of a human hair). Therefore, for the first time this study will use the new generation of high resolution Secondary Ion Mass Spectrometers permitting the measurement of fine-scale chemical zonation of plagioclase crystals from Mount St. Helen's in the USA and Mount Taranaki in New Zealand. This will allow the magmatic processes and the timescales over which these processes occurred directly prior to eruption to be assessed. These timescales can then be evaluated against the known timescales for the movement of magma prior to the recent eruptions of Mount St. Helen's allowing better models for the future prediction of volcanic eruptions to be constrained. This will provide valuable insights into the working of volcanos directly prior to eruption and could have immense benefits to the millions of people worldwide who live in the shadow of volcanoes and help mitigate the associated hazards of active volcanoes. This research will be conducted in the Department of Earth Sciences, University of Bristol in conjunction with the Interface Analysis Centre, Bristol, Muenster University, Germany, University of Western Australia, Cascades Volcano Observatory and University of Orogen in the USA and Victoria University of Wellington, New Zealand.