Compositional variations of ignimbrite magmas in the Central Andes over the past 26 Ma — A multivariate statistical perspective
Compositional variations of ignimbrite magmas in the Central Andes over the past 26 Ma — A multivariate statistical perspective
复制标题
过去 26 Ma 中安第斯山脉火凝结岩浆的成分变化 â 多元统计视角
DOI：
10.1016/j.lithos.2016.07.011
复制
发表时间：
2016
期刊：
影响因子：
3.5
通讯作者：
Wörner G
中科院分区：
文献类型：
--
作者：
Brandmeier M;Wörner G
多元统计和地理空间分析基于来自安第斯山脉中部 194 个映射的火凝灰岩的 890 个地球化学数据和约 1200 个地质年代学数据的汇编，记录了新近纪时期大量火凝灰岩（所谓的“火凝灰岩爆发”）的成分和时间模式。过去十年中计算科学的快速发展导致针对具有许多预测变量的大型数据集的多元统计算法池不断增长。本研究对对数比转换数据应用聚类分析 (CA) 和线性判别分析 (LDA)，目的是 (1) 测试熔结岩相关性工具，以及 (2) 区分反映安第斯山脉中部地球动力学演化过程中不同过程和熔结岩岩浆作用来源的成分组。主要元素和微量元素的 CA 允许根据其地球化学特征将熔结岩分组为流纹岩和英安岩“端元”，并区分有关 Eu 异常、中重稀土元素 (REE) 的损耗以及轻 REE 的可变富集的特征微量元素特征。为了突出这些独特的成分特征，我们将 LDA 应用于选定的具有综合数据集的熔结岩。与传统的地球化学参数相比，我们发现多元统计的优点是能够处理大型数据集和许多变量（元素），并利用这个n维空间来检测数据中包含的细微成分差异。区分熔凝灰岩最重要的预测因子是La、Yb、Eu、Al2O3、K2O、P2O5、MgO、FeOt和TiO2。然而，其他稀土元素，如 Gd、Pr、Tm、Sm、Dy 和 Er 也对区分函数有贡献。在 (1) 智利最北部和秘鲁南部较古老 (> 13 Ma) 的大体积高原形成熔结岩和 (2) 体积相似的较年轻 (< 10 Ma) 高原-普纳火山杂岩 (APVC) 熔结岩之间发现了显着的成分差异。较老的凝结岩中重稀土元素的消耗较少，锶同位素的放射性也较少，这表明在较薄且热演化程度较低的地壳中，地壳在演化过程中的贡献较小。这些成分变化表明与地壳增厚有关，从斜长石到角闪石和石榴石残余矿物的“转变”在 13 Ma 到 9 Ma 之间。成分和体积变化与胡安-费尔南德斯海岭的南北走向、地壳缩短和增厚以及过去 26 Ma 期间平均地壳温度升高相关。
Multivariate statistical and geospatial analyses based on a compilation of 890 geochemical and ~ 1200 geochronological data for 194 mapped ignimbrites from the Central Andes document the compositional and temporal patterns of large-volume ignimbrites (so-called “ignimbrite flare-ups”) during Neogene times. Rapid advances in computational science during the past decade led to a growing pool of algorithms for multivariate statistics for large datasets with many predictor variables. This study applies cluster analysis (CA) and linear discriminant analysis (LDA) on log-ratio transformed data with the aim of (1) testing a tool for ignimbrite correlation and (2) distinguishing compositional groups that reflect different processes and sources of ignimbrite magmatism during the geodynamic evolution of the Central Andes. CA on major and trace elements allows grouping of ignimbrites according to their geochemical characteristics into rhyolitic and dacitic “end-members” and to differentiate characteristic trace element signatures with respect to Eu anomaly, depletions in middle and heavy rare earth elements (REE) and variable enrichments in light REE. To highlight these distinct compositional signatures, we applied LDA to selected ignimbrites for which comprehensive datasets were available. In comparison to traditional geochemical parameters we found that the advantage of multivariate statistics is their capability of dealing with large datasets and many variables (elements) and to take advantage of this n-dimensional space to detect subtle compositional differences contained in the data.The most important predictors for discriminating ignimbrites are La, Yb, Eu, Al2O3, K2O, P2O5, MgO, FeOt,and TiO2. However, other REE such as Gd, Pr, Tm, Sm, Dy and Er also contribute to the discrimination functions.Significant compositional differences were found between (1) the older (> 13 Ma) large-volume plateau-forming ignimbrites in northernmost Chile and southern Peru and (2) the younger (< 10 Ma) Altiplano–Puna–Volcanic-Complex (APVC) ignimbrites that are of similar volumes. Older ignimbrites are less depleted in HREE and less radiogenic in Sr isotopes, indicating smaller crustal contributions during evolution in a thinner and thermally less evolved crust. These compositional variations indicate a relation to crustal thickening with a “transition” from plagioclase to amphibole and garnet residual mineralogy between 13 and 9 Ma. Compositional and volumetric variations correlate to the N–S passage of the Juan-Fernandéz-Ridge, crustal shortening and thickening, and increased average crustal temperatures during the past 26 Ma.