Spatial heterogeneity analysis of potentially toxic elements for mapping environmental risk areas in soil

Potentially Toxic Elements (PTEs) are a critical focus of soil environmental management. Proactively identifying risks associated with PTEs is a key research focus. This study utilized the spatial variation coefficient of soil PTEs, contrasting traditional pollution evaluation methods. By integrating outlier detection and scale transformation, we aimed to create an automated risk zone identification and mapping process, analyzing how outlier determination conditions and spatial resolution affect risk zone mapping. The results indicate: (1) Strict determination conditions (Z-scores>2) can accurately identify high-risk zones but at the expense of potential risk zones (accuracy rate of only 6 %). In contrast, lenient determination conditions (Z-scores>3) distribute identification errors across different risk levels, which are detrimental to the identification and mapping of high-risk zones. (2) Higher spatial resolution improves the precision of the identified risk zones; however, a balance must be established between mapping accuracy and computational speed. To this end, this study proposes a decay index to quantify the relationship between the resolution and high-risk zone identification.