Role of cryptic amphibole crystallization in magma differentiation at Hudson volcano, Southern Volcanic Zone, Chile
Role of cryptic amphibole crystallization in magma differentiation at Hudson volcano, Southern Volcanic Zone, Chile
复制标题
DOI：
10.1007/s00410-009-0426-1
复制
发表时间：
2010-02-01
影响因子：
3.5
通讯作者：
Naranjo, Jose-Antonio
中科院分区：
文献类型：
--
作者：
Kratzmann, David J.;Carey, Steven;Naranjo, Jose-Antonio
哈德逊火山（智利）是安第斯山脉南部火山带最南端的成层火山，曾发生过南美洲一些最大的全新世火山喷发。哈德逊至少发生过 12 次全新世爆发事件，其中距今 6700 年、距今 3600 年和 1991 年的喷发次数最多。哈得逊火山一直喷出具有相似粗面岩质和粗面英安岩成分的岩浆，具有相似的无水斑晶组合、喷发前温度和氧逸度。三个最大的全新世事件的喷发前储存条件已经通过矿物地温测量、熔体包裹体挥发物含量以及与类似高压实验的比较进行了估计。在整个全新世，粗岩岩浆的储存发生在 0.2 至 2.7 km 的深度，温度大约为 972A 摄氏度 (+/- 25)，log fO(2) -10.33-10.24 (+/- 0.2)（比 NNO 缓冲液高一个 log 单位），熔体中含有 1 至 3 wt% H2O。粗面英质岩浆喷发前的储存发生在 1.1 至 2.0 km 之间，温度类似于 942A 摄氏度 (+/- 26) 和 log fO(2) -10.68 (+/- 0.2)，熔体中的水含量类似于 2.5 wt% H2O。演化的粗面质岩浆和粗面英质质岩浆可主要通过分级结晶源自玄武岩母体。根据斜长石熔体包裹体估计的截留压力表明结晶水平相对较浅。然而，微量元素数据（例如 Dy/Yb 比率趋势）表明角闪石在哈德逊岩浆的分异中发挥了重要作用，并且这种分馏可能发生在 > 6 km 的深度。 Hudson 微量元素数据中缺乏石榴石信号，母体镁铁质岩浆分化的潜在分期点 [即，类似于 20 km（例如，Annen 等人，在 J Petrol 47(3):505-539, 2006）]，以及推断的角闪岩相 [类似于 24 km（例如，Rudnick 和 Fountain） Rev Geophys 33:267-309, 1995）]结合起来对分异深度的下限（即，类似于 20-24 km）施加一些限制。这些限制表明，地幔来源的岩浆分异发生在中上至下地壳水平，涉及包括角闪石的含水矿物组合，并产生类似于 1991 年喷发第一阶段喷出的岩浆的玄武岩到玄武岩安山岩成分。在此深度继续分馏导致粗面岩和粗面英质岩组合物的形成。这些更进化的岩浆在浅地壳中上升并停滞，正如从熔体包裹体获得的俘获压力所表明的那样。伴随上升的压力下降，加上减压引起的结晶对岩浆体的潜在加热，将导致岩浆穿过角闪石稳定场。进一步的浅层结晶涉及无水矿物组合，这可以解释哈得逊岩套中缺乏斑晶角闪石。
Hudson volcano (Chile) is the southern most stratovolcano of the Andean Southern Volcanic Zone and has produced some of the largest Holocene eruptions in South America. There have been at least 12 recorded Holocene explosive events at Hudson, with the 6700 years BP, 3600 years BP, and 1991 eruptions the largest of these. Hudson volcano has consistently discharged magmas of similar trachyandesitic and trachydacitic composition, with comparable anhydrous phenocryst assemblages, and pre-eruptive temperatures and oxygen fugacities. Pre-eruptive storage conditions for the three largest Holocene events have been estimated using mineral geothermometry, melt inclusion volatile contents, and comparisons to analogous high pressure experiments. Throughout the Holocene, storage of the trachyandesitic magmas occurred at depths between 0.2 and 2.7 km at approximately similar to 972A degrees C (+/- 25) and log fO(2) -10.33-10.24 (+/- 0.2) (one log unit above the NNO buffer), with between 1 and 3 wt% H2O in the melt. Pre-eruptive storage of the trachydacitic magma occurred between 1.1 and 2.0 km, at similar to 942A degrees C (+/- 26) and log fO(2) -10.68 (+/- 0.2), with similar to 2.5 wt% H2O in the melt. The evolved trachyandesitic and trachydacitic magmas can be derived from a basaltic parent primarily via fractional crystallization. Entrapment pressures estimated from plagioclase-hosted melt inclusions suggest relatively shallow levels of crystallization. However, trace element data (e.g., Dy/Yb ratio trends) suggests amphibole played an important role in the differentiation of the Hudson magmas, and this fractionation is likely to have occurred at depths > 6 km. The absence of a garnet signal in the Hudson trace element data, the potential staging point for differentiation of parental mafic magmas [i.e., similar to 20 km (e.g., Annen et al. in J Petrol 47(3):505-539, 2006)], and the inferred amphibolite facies [similar to 24 km (e.g., Rudnick and Fountain in Rev Geophys 33:267-309, 1995)] combine to place some constraint on the lower limit of depth of differentiation (i.e., similar to 20-24 km). These constraints suggest that differentiation of mantle-derived magmas occurred at upper-mid to lower crustal levels and involved a hydrous mineral assemblage that included amphibole, and generated a basaltic to basaltic andesitic composition similar to the magma discharged during the first phase of the 1991 eruption. Continued fractionation at this depth resulted in the formation of the trachyandesitic and trachydacitic compositions. These more evolved magmas ascended and stalled in the shallow crust, as suggested by the pressures of entrapment obtained from the melt inclusions. The decrease in pressure that accompanied ascent, combined with the potential heating of the magma body through decompression-induced crystallization would cause the magma to cross out of the amphibole stability field. Further shallow crystallization involved an anhydrous mineral assemblage and may explain the lack of phenocrystic amphibole in the Hudson suite.