Katseanes Chelsea K, Chappell Mark A, Hopkins Bryan G, Durham Brian D, Price Cynthia L, Porter Beth E, Miller Lesley F
Department of Plant and Wildlife Sciences, Brigham Young University, Provo, UT, USA.
Environmental Laboratory, US Army Engineer Research & Development Center, Vicksburg, MS, USA.
J Environ Manage. 2016 Nov 1;182:101-110. doi: 10.1016/j.jenvman.2016.07.043. Epub 2016 Jul 22.
After nearly a century of use in numerous munition platforms, TNT and RDX contamination has turned up largely in the environment due to ammunition manufacturing or as part of releases from low-order detonations during training activities. Although the basic knowledge governing the environmental fate of TNT and RDX are known, accurate predictions of TNT and RDX persistence in soil remain elusive, particularly given the universal heterogeneity of pedomorphic soil types. In this work, we proposed a new solution for modeling the sorption and persistence of these munition constituents as multivariate mathematical functions correlating soil attribute data over a variety of taxonomically distinct soil types to contaminant behavior, instead of a single constant or parameter of a specific absolute value. To test this idea, we conducted experiments measuring the sorption of TNT and RDX on taxonomically different soil types that were extensively physical and chemically characterized. Statistical decomposition of the log-transformed, and auto-scaled soil characterization data using the dimension-reduction technique PCA (principal component analysis) revealed a strong latent structure based in the multiple pairwise correlations among the soil properties. TNT and RDX sorption partitioning coefficients (KD-TNT and KD-RDX) were regressed against this latent structure using partial least squares regression (PLSR), generating a 3-factor, multivariate linear functions. Here, PLSR models predicted KD-TNT and KD-RDX values based on attributes contributing to endogenous alkaline/calcareous and soil fertility criteria, respectively, exhibited among the different soil types: We hypothesized that the latent structure arising from the strong covariance of full multivariate geochemical matrix describing taxonomically distinguished soil types may provide the means for potentially predicting complex phenomena in soils. The development of predictive multivariate models tuned to a local soil's taxonomic designation would have direct benefit to military range managers seeking to anticipate the environmental risks of training activities on impact sites.
在众多弹药平台上使用了近一个世纪后，由于弹药制造或作为训练活动中低阶爆炸释放物的一部分，三硝基甲苯（TNT）和黑索金（RDX）污染物大量出现在环境中。尽管关于TNT和RDX在环境中归宿的基本知识已为人所知，但要准确预测TNT和RDX在土壤中的持久性仍然困难重重，尤其是考虑到土壤类型普遍存在的非均质性。在这项研究中，我们提出了一种新的解决方案，将这些弹药成分的吸附和持久性建模为多元数学函数，该函数将各种分类学上不同的土壤类型的土壤属性数据与污染物行为相关联，而不是单一常数或特定绝对值的参数。为了验证这一想法，我们进行了实验，测量TNT和RDX在分类学上不同的土壤类型上的吸附情况，这些土壤类型在物理和化学方面都有广泛的特征描述。使用降维技术主成分分析（PCA）对经对数转换和自动缩放的土壤特征数据进行统计分解，揭示了基于土壤属性之间多重成对相关性的强大潜在结构。使用偏最小二乘回归（PLSR）将TNT和RDX的吸附分配系数（KD-TNT和KD-RDX）与这种潜在结构进行回归，生成了一个三因素多元线性函数。在此，PLSR模型分别根据不同土壤类型中对内生碱性/钙质和土壤肥力标准有贡献的属性预测KD-TNT和KD-RDX值：我们假设，描述分类学上不同土壤类型的全多元地球化学矩阵的强协方差产生的潜在结构可能为潜在预测土壤中的复杂现象提供手段。针对当地土壤分类指定开发的预测性多元模型将直接有利于军事靶场管理人员，他们试图预测训练活动对撞击场地的环境风险。
