Rapid assessment of soil contamination by potentially toxic metals in the green spaces of Moscow megalopolis using the portable X-ray analyzer

Abstract

Purpose

Anthropogenic influence leads to significant changes in soil properties and functions. Soil contamination by potentially toxic metals is one of the major environmental problems in urban environments. Traditional soil monitoring methods, while accurate, are often costly and labor-intensive, making it challenging to capture the intricate spatial variations of pollutants in urban soils. Proximal sensing based on X-ray fluorescence (XRF) analysis is considered a cost-effective approach for rapid assessment of soil contamination. The assessment accuracy depends on soil properties (e.g., texture, moisture, organic matter content) and detection limits for different elements. The research aimed to test a portable XRF analyzer for the assessment of soil contamination by potentially toxic metals in green zones of Moscow megalopolis.

Materials and methods

Initially, Olympus Vanta C pXRF was calibrated using artificially contaminated soil mixtures by Ni, Cu, Pb, Zn, and Cd, representing a diversity of urban soils in Moscow. Linear regression was used to compare pXRF results with the ICP-OES method, and regression coefficients were used to set correction factors (k) for observed potentially toxic metals based on soil properties. Subsequently, the spatial mapping accuracy of topsoil contamination in three distinct green areas was assessed using pXRF (with and without correction factors) based on ICP-OES reference concentrations.

Results

The calibrated pXRF showed high accuracy for Pb (R² = 0.94, b = 0.91, p < 0.05), Cu (R² = 0.95, b = 0.95, p < 0.05), and Zn (R² = 0.95, b = 1.04, p < 0.05), moderate accuracy for Ni (R² = 0.68, b = 0.77, p < 0.05), and limited accuracy for Cd (p > 0.05) on a typical urban contamination level due to its high detection limit. Spatial variability in soil contamination was determined by comparison to the health thresholds (approximate permissible concentration and pollution indices), and the areas subjected to land-use restrictions were identified based on the regional environmental regulations. When calibrated by correction factors, mapping accuracy based on pXRF approached that of ICP-OES (in the range of 10%) for Ni, Cu, and Pb in major parts of the areas.

Conclusion

The study revealed that uncorrected pXRF measurements overestimated contamination. When tailored to specific urban soil conditions, pXRF offers a viable, cost-efficient alternative for assessing soil contamination. The developed approach improved the accuracy and reliability of local soil contamination maps by capturing spatial patterns ignored by conventional methods which is essential to optimize costs of soil rehabilitation and sustainable management of urban soils.