Over the last decade, a large number of studies have been conducted on heavy metals and magnetic susceptibility (${\mathsf{\chi }}_{\mathrm{l}\mathrm{f}}$) measurement in soils. Yet, a global understanding of soil contamination and magnetic responses remains elusive due to the limited scope or sampling sites of these studies. Hence, we attempted to explore a pollution proxy on a global scale. Through a meta-analysis of data from 102 published studies, our research aimed to provide a worldwide overview of heavy metal pollution and magnetic responses in agriculture soils. We mapped the geographic distribution of nine heavy metals (Cr, Cu, Zn, Pb, Ni, As, Cd, Mn, and Fe) in agricultural soils and explored their pollution sources and contributions. Since 2011, The accumulation of heavy metals has escalated, with industrial activities (31.5%) being the largest contributor, followed by agricultural inputs (27.1%), atmospheric deposition (22.66%), and natural sources (18.74%). The study reports ${\mathsf{\chi }}_{\mathrm{l}\mathrm{f}}$ ranging from 6.45 × 10⁻⁸ m³/kg to 319.23 × 10⁻⁸ m³/kg and ${\mathsf{\chi }}_{\mathrm{f}\mathrm{d}}$ from 0.59% and 12.85%, with the majority of the samples being below 6%, indicating heavy metal influence mainly from human activities. Pearson’s correlation and redundancy analysis show significant positive correlations of Pb, Zn, and Cu with ${\mathsf{\chi }}_{\mathrm{l}\mathrm{f}}$ (r = 0.51–0.53) and Mn and Fe with ${\mathsf{\chi }}_{\mathrm{f}\mathrm{d}}$ (r = 0.50–0.53), while Pb, Zn, Cu, and As metals were shown to be key factors of variation in magnetic response. The average heavy metal pollution load index of 2.03 suggests moderate global agricultural soil pollution, with higher heavy metal contamination in areas of high ${\mathsf{\chi }}_{\mathrm{l}\mathrm{f}}$. Regression analysis confirms soil is considered to be non-polluted below ${\mathsf{\chi }}_{\mathrm{l}\mathrm{f}}$ of $26\times {10}^{-8} {\mathrm{m}}^3/\mathrm{k}\mathrm{g}$ and polluted above this threshold, with all contamination factors of metals showing a linear correlation with ${\mathsf{\chi }}_{\mathrm{l}\mathrm{f}}$ (R = 0.72), indicating that a significant relationship between ${\mathsf{\chi }}_{\mathrm{l}\mathrm{f}}$ and the geochemical properties of soils continues to exist on a global scale. This study provides new insights for large-scale agricultural soil quality assessment and magnetic response.

Soil fertility is a critical factor in agricultural production, directly impacting crop growth, yield, and quality. To achieve precise agricultural management, accurate spatial interpolation of soil parameters is essential. This study developed a new interpolation prediction framework that combines Radial Basis Function (RBF) neural networks with Inverse Distance Weighting (IDW), termed the IDW-RBFNN. This framework initially uses the IDW method to apply preliminary weights based on distance to the data points, which are then used as input for the RBF neural network to form a training dataset. Subsequently, the RBF neural network further trains on these data to refine the interpolation results, achieving more precise spatial data interpolation. We compared the interpolation prediction accuracy of the IDW-RBFNN framework with ordinary Kriging (OK) and RBF methods under three different parameter settings. Ultimately, the IDW-RBFNN demonstrated lower error rates in terms of RMSE and MRE compared to direct RBF interpolation methods when adjusting settings based on different power values, even with a fixed number of data samples. As the sample size decreases, the interpolation accuracy of OK and RBF methods is significantly affected, while the error of IDW-RBFNN remains relatively low. Considering both interpolation accuracy and resource limitations, we recommend using the IDW-RBFNN method (p = 2) with at least 60 samples as the minimum sampling density to ensure high interpolation accuracy under resource constraints. Our method overcomes limitations of existing approaches that use fixed steady-state distance decay parameters, providing an effective tool for soil fertility monitoring in delta regions.