The understanding of sources and controlling factors of potentially toxic elements (PTEs) in soils plays an important role in the improvement of environmental management. With the rapid growth of data volume, effective methods are required for data analytics for the large geochemical data sets. In recent years, spatial machine learning technologies have been proven to have the potential to reveal hidden spatial patterns in order to extract geochemical information. In this study, two spatial clustering techniques of Getis-Ord Gi* statistic and K-means clustering analysis were performed on 15 PTEs in 6,862 topsoil samples from the Tellus datasets of Northern Ireland to investigate the hidden spatial patterns and association with their controlling factors. The spatial clustering patterns of hot spots (high values) and cold spots (low values) for the 15 PTEs were revealed, showing clear association with geological features, especially peat and basalt. Peat was associated with high concentrations of Bi, Pb, Sb and Sn, while basalt was associated with high concentrations of Co, Cr, Cu, Mn, Ni, V and Zn. The high concentrations of As, Ba, Mo and U were associated with mixture of various lithologies, indicating the complicated influences on them. In addition, three hidden patterns in the 6,862 soil samples were revealed by K-means clustering analysis. The soil samples in the first and second clusters were overlaid on the peatland and basalt formation, respectively, while the samples in the third cluster were overlaid on the mixture of the other lithologies. These hidden patterns of soil samples were consistent with the spatial clustering patterns for PTEs, highlighting the dominant control of peat and basalt in the topsoil of Northern Ireland. This study demonstrates the power of spatial machine learning techniques in identifying hidden spatial patterns, providing evidences to extract geochemical knowledge in environmental studies.

Total organic carbon (TOC) has received increased attention in recent years, not only as an important indicator in soil fertility, but also due to its close relationship with the atmosphere. Generally, soil TOC and pH values follow a negative correlation, which was revealed by traditional statistical methods. However, the conventional global models lack the ability to capture the spatial variation locally. In this study, spatially varying local relationships between TOC and pH values are studied by geographically weighted regression (GWR) on continental-scale data of European agricultural soil from the project 'Geochemical Mapping of Agricultural and Grazing land Soil' (GEMAS). In this study, TOC is the dependent and pH the independent variable. Both negative and positive local correlation coefficients are observed, showing the existence of 'special' spatially varying relationships between TOC and pH values. Original negative relationships change to positive values in more than 50% of the study area. Novel finding of significant positive correlations is observed in central-eastern Europe, while negative correlations are found mainly in northern Europe. Mixed relationships occur in southern Europe. These special patterns are strongly associated with specific natural factors, especially the extensive occurrence of quartz-rich soil in the central-eastern part of Europe. Anthropogenic inputs may have also played a role in the mixed southern European areas. The GWR technique is powerful and effective for revealing spatially varying relationships at the local level. Thus, it provides a new way to further explore the related influencing factors on the TOC and pH spatial distribution.

Potentially toxic elements (PTEs) and polycyclic aromatic hydrocarbons (PAHs) harm the ecosystem and human health, especially in urban areas. Identifying and understanding their potential sources and underlying interactions in urban soils are critical for informed management and risk assessment. This study investigated the potential sources and the spatially varying relationships between 9 PTEs and PAHs in the topsoil of Dublin by combining positive matrix factorisation (PMF) and geographically weighted regression (GWR). The PMF model allocated four possible sources based on species concentrations and uncertainties. The factor profiles indicated the associations with high-temperature combustion (PAHs), natural lithologic factors (As, Cd, Co, Cr, Ni), mineralisation and mining (Zn), as well as anthropogenic inputs (Cu, Hg, Pb), respectively. In addition, selected representative elements Cr, Zn, and Pb showed distinct spatial interactions with PAHs in the GWR model. Negative relationships between PAHs and Cr were observed in all samples, suggesting the control of Cr concentrations by natural factors. Negative relationships between PAHs and Zn in the eastern and north-eastern regions were related to mineralisation and anthropogenic Zn-Pb mining. In contrast, the surrounding regions exhibited a natural relationship between these two variables with positive coefficients. Increasing positive coefficients from west to east were observed between PAHs and Pb in the study area. This special pattern was consistent with prevailing south-westerly wind direction in Dublin, highlighting the predominant influences on PAHs and Pb concentrations from vehicle and coal combustion through atmospheric deposition. Our results provided a better understanding of geochemical features for PTEs and PAHs in the topsoil of Dublin, demonstrating the efficiency of combined approaches of receptor models and spatial analysis in environmental studies.

Peri-urban agricultural soils are often contaminated by potentially toxic elements (PTEs) due to rapid urbanization, industrial activities, and agricultural practices. In this study, two advanced analytical methods including positive matrix factorization (PMF) model and K-means clustering algorithm were integrated to explore the potential sources and concealed contamination patterns of 8 PTEs in peri-urban soils in county Gaoming, China. Descriptive statistics showed average concentrations of arsenic (As), cadmium (Cd), chromium (Cr), copper (Cu), mercury (Hg), nickel (Ni), lead (Pb), and zinc (Zn) as 19.11, 0.18, 35.69, 20.31, 18.26, 151.7, 67.75, and 0.29 mg/kg, respectively. The PMF model identified three primary sources: geogenic (Cr, Ni), industrial and traffic-related (Pb, Hg, Zn), and agricultural (As, Cd and Cu). The contribution of each source was quantified: geogenic sources contributed 55.6% to Cr and 52.3% to Ni, industrial sources accounted for 41.8% of Pb, 58.4% of Hg, and 41.9% of Zn, while agricultural practices contributed 88.1% of As, 77.9% of Cu, and 70.7% of Cd. Subsequently, K-means clustering classified the soil samples into three distinct clusters based on the derived factor contribution from PMF model, reflecting their clear spatial associations with different types of land use: large-scale agricultural areas (Cluster 1), natural vegetation (Cluster 2), and urbanized zones (Cluster 3). Furthermore, boxplots showed that the highest PTE concentrations were found in the third cluster, confirming the significant impact of human activities, while the lower concentrations in the second cluster indicated more natural conditions. These results underscored the dual influences of agriculture and urbanization on PTE contamination, which highlighted the need for targeted soil management strategies. Moreover, the integration of PMF and K-means clustering effectively reveals potential sources and concealed pollution patterns, providing insights for managing pollution and safeguarding environmental health in rapidly urbanized areas.