School of Engineering, RMIT University, GPO Box 2476, Melbourne, Victoria 3001, Australia; Centre for Environmental Sustainability and Remediation, RMIT University, Victoria, Australia.
Centre for Environmental Sustainability and Remediation, RMIT University, Victoria, Australia; School of Science, RMIT University, Victoria, Australia.
Sci Total Environ. 2017 Feb 15;580:1410-1420. doi: 10.1016/j.scitotenv.2016.12.106. Epub 2016 Dec 23.
Understanding ambient background concentrations in soil, at a local scale, is an essential part of environmental risk assessment. Where high resolution geochemical soil surveys have not been undertaken, soil data from alternative sources, such as environmental site assessment reports, can be used to support an understanding of ambient background conditions. Concentrations of metals/metalloids (As, Mn, Ni, Pb and Zn) were extracted from open-source environmental site assessment reports, for soils derived from the Newer Volcanics basalt, of Melbourne, Victoria, Australia. A manual screening method was applied to remove samples that were indicated to be contaminated by point sources and hence not representative of ambient background conditions. The manual screening approach was validated by comparison to data from a targeted background soil survey. Statistical methods for exclusion of contaminated samples from background soil datasets were compared to the manual screening method. The statistical methods tested included the Median plus Two Median Absolute Deviations, the upper whisker of a normal and log transformed Tukey boxplot, the point of inflection on a cumulative frequency plot and the 95th percentile. We have demonstrated that where anomalous sample results cannot be screened using site information, the Median plus Two Median Absolute Deviations is a conservative method for derivation of ambient background upper concentration limits (i.e. expected maximums). The upper whisker of a boxplot and the point of inflection on a cumulative frequency plot, were also considered adequate methods for deriving ambient background upper concentration limits, where the percentage of contaminated samples is <25%. Median ambient background concentrations of metals/metalloids in the Newer Volcanic soils of Melbourne were comparable to ambient background concentrations in Europe and the United States, except for Ni, which was naturally enriched in the basalt-derived soils of Melbourne.
了解土壤中的环境背景浓度是环境风险评估的重要组成部分。在没有进行高分辨率地球化学土壤调查的情况下，可以使用替代来源（如环境场地评估报告）的土壤数据来支持对环境背景条件的理解。从澳大利亚维多利亚州墨尔本的新型火山玄武岩衍生的土壤中，从公开的环境场地评估报告中提取了金属/类金属（As、Mn、Ni、Pb 和 Zn）的浓度。应用手动筛选方法去除了表明受到点源污染且因此不能代表环境背景条件的样本。通过与目标背景土壤调查的数据进行比较，验证了手动筛选方法的有效性。用于从背景土壤数据集中排除污染样本的统计方法与手动筛选方法进行了比较。测试的统计方法包括中位数加两个中位数绝对偏差、正态分布和对数转换的 Tukey 箱线图的上须、累积频率图上的拐点以及第 95 百分位数。我们已经证明，在无法使用现场信息筛选异常样本结果的情况下，中位数加两个中位数绝对偏差是推导环境背景上限浓度（即预期最大值）的保守方法。箱线图的上须和累积频率图上的拐点也是推导环境背景上限浓度的合适方法，前提是污染样本的百分比<25%。墨尔本新型火山土壤中金属/类金属的中位数环境背景浓度与欧洲和美国的环境背景浓度相当，除了 Ni，它在墨尔本的玄武岩衍生土壤中自然富集。
