The geological imprint on the geochemical background in stream sediment of the Itacaiúnas River Watershed (IRW), located in the most prominent mining area of Brazil, the Carajás Mineral Province, was evaluated. The stream sediment dataset from the Itacaiúnas Geochemical Mapping and Background Project of the Instituto Tecnológico Vale was used to address multi-element associations and for the definition of surface geochemical compartments by applying a series of multivariate geostatistics tools and geoprocessing techniques, through the Compositional Data Analysis (CoDA) perspective. In this study, the geochemical dataset was processed for centered log-ratio (clr) transformation prior to the Principal Factor Analysis (PFA). The spatial distribution of factor scores derived from PFA indicates a pronounced control on stream sediment geochemistry by the underlying geological setting and lithology. Relevant hidden correlations were observed among the nine extracted factors. An unsupervised classification based on k-means clustering algorithm was performed for the factor scores of 761 stream sediment samples. A multiproxy approach that combines the cluster classification and geoprocessing tools revealed a strong similarity in the spatial distribution of different geochemical compartments and geological domains of the IRW. Threshold values of 51 elements were calculated by a variety of statistical methods for references to the main four geological domains of the IRW defined using the geological framework and the geochemical compartments. These values were then compared to the background concentrations of the entire study area and to the investigation levels proposed by the Brazilian environmental agencies for potentially toxic elements. The new threshold values for the geochemical compartments of the IRW can be used for improving the assessment of environmental impact by anthropic activities, and to establish more realistic threshold values for different chemical elements in stream sediments. This study provides useful information for new mineral development projects and ore prospects, with potential benefit to the mineral sector and industry.

对位于巴西最著名的采矿区卡拉哈斯矿产省的伊塔卡尤纳斯河流域（IRW）的河流沉积物在地球化学背景上的地质印记进行了评估。来自TecnológicoVale研究所的Itacaiúnas地球化学测绘和背景项目的河流沉积物数据集，通过成分数据分析，通过应用一系列多元地统计学工具和地理处理技术，用于解决多种元素的关联并定义地表地球化学区室。 （CoDA）角度。在这项研究中，在进行主因子分析（PFA）之前，先对地球化学数据集进行中心对数比（clr）转换。从PFA得出的因子得分的空间分布表明，潜在的地质环境和岩性对河流沉积物地球化学有明显的控制作用。在九个提取因子之间观察到相关的隐藏相关性。基于761个河流沉积物样本的因子得分，进行了基于k均值聚类算法的无监督分类。结合聚类分类和地理处理工具的多代理方法显示了IRW不同地球化学区室和地质域在空间分布上的强烈相似性。通过多种统计方法计算了51个元素的阈值，以参考使用地质框架和地球化学隔室定义的IRW的主要四个地质域。然后将这些值与整个研究区域的背景浓度以及巴西环境机构针对潜在有毒元素提出的调查水平进行比较。IRW的地球化学隔室的新阈值可用于改善人类活动对环境影响的评估，并为河流沉积物中的不同化学元素建立更实际的阈值。这项研究为新的矿产开发项目和矿产前景提供了有用的信息，对矿产部门和工业都有潜在的好处。IRW的地球化学隔室的新阈值可用于改善人类活动对环境影响的评估，并为河流沉积物中的不同化学元素建立更实际的阈值。这项研究为新的矿产开发项目和矿产前景提供了有用的信息，对矿产部门和工业都有潜在的好处。IRW的地球化学隔室的新阈值可用于改善人类活动对环境影响的评估，并为河流沉积物中的不同化学元素建立更实际的阈值。这项研究为新的矿产开发项目和矿产前景提供了有用的信息，对矿产部门和工业都有潜在的好处。