Improved enrichment factor calculations through principal component analysis: Examples from soils near breccia pipe uranium mines, Arizona, USA
Improved enrichment factor calculations through principal component analysis: Examples from soils near breccia pipe uranium mines, Arizona, USA
复制标题
DOI：
10.1016/j.envpol.2019.01.122
复制
发表时间：
2019-05-01
影响因子：
8.9
通讯作者：
Naftz, David L.
中科院分区：
文献类型：
--
作者：
Bern, Carleton R.;Walton-Day, Katie;Naftz, David L.
富集因子 (EF) 是一种广泛使用的指标，用于确定采样介质中元素的存在相对于平均自然丰度由于人类活动而增加了多少。 EF 的计算需要选择背景成分和参考元素，这些选择会强烈影响计算结果。这里展示了如何通过双图仔细应用经典主成分分析 (PCA) 来指导背景成分和参考元素的选择。使用中心对数比 (CLR) 变换处理元素数据，并检查主要元素和微量元素的多个子集以获得不同的视角。该方法适用于美国亚利桑那州角砾岩管铀矿周围元素土壤浓度的数据集，大多数样本是通过增量采样方法收集的。矿石在地表的储存可能会导致矿石衍生材料被风吹散。铀被发现是矿石衍生物质扩散到附近土壤的最佳示踪剂，EF 值高达 75。硫、砷、钼和铜也被富集，但程度较低。结果证明了 PCA 在这些情况下的几个实际好处：（1）能够识别最适合区分特定富集源和背景成分的一个或多个元素； (2) 了解站点内部和站点之间的背景成分如何变化； (3) 根据其综合的多元素成分鉴定含有富集或人为材料的样品。计算最具代表性的 EF 值对于富集的数值评估非常有用，无论是人为的还是自然的。然而，如此处所示，PCA 和双标图方法提供了一种可视化方法，可以集成给定数据子集的所有元素的信息，从而产生超出 EF 能力的地球化学见解。由 Elsevier Ltd 出版。这是一篇基于 CC BY 许可证 (http://creativecommons.org/licenses/by/4.0/) 的开放获取文章。
The enrichment factor (EF) is a widely used metric for determining how much the presence of an element in a sampling media has increased relative to average natural abundance because of human activity. Calculation of an EF requires the selection of both a background composition and a reference element, choices that can strongly influence the result of the calculation. Here, it is shown how carefully applied, classical principal component analysis (PCA) examined via biplots can guide the selections of background compositions and reference elements. Elemental data were treated using the centered log ratio (CLR) transformation, and multiple subsets of major and trace elements were examined to gain different perspectives. The methodology was applied to a dataset of elemental soil concentrations from around breccia pipe uranium mines in Arizona, U.S.A., with most samples collected via incremental sampling methodology. Storage of ore at the surface creates the potential for wind dispersal of ore derived material. Uranium was found to be the best individual tracer of dispersal of ore-derived material to nearby soils, with EF values up to 75. Sulfur, As, Mo, and Cu were also enriched but to lesser degrees. The results demonstrate several practical benefits of a PCA in these situations: (1) the ability to identify one or more elements best suited to distinguish a specific source of enrichment from background composition; (2) understanding how background compositions vary within and between sites; (3) identification of samples containing enriched or anthropogenic materials based upon their integrated, multi-element composition. Calculating the most representative EF values is useful for numerical assessment of enrichment, whether anthropogenic or natural. As shown here, however, the PCA and biplot method provide a visual approach that integrates information from all elements for a given subset of data in a manner that yields geochemical insights beyond the power of the EF. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).