Related Experiment Video
Updated: May 26, 2025
Use of Principal Components for Scaling Up Topographic Models to Map Soil Redistribution and Soil Organic Carbon
Published on: October 16, 2018
Uncovering soil heavy metal pollution hotspots and influencing mechanisms through machine learning and spatial
Xiaoyong Song1, Yao Sun2, Huijuan Wang3
1Department of Environmental Science and Engineering, Fudan University, Shanghai, 200433, China; Shanghai Academy of Landscape Architecture Science and Planning, Shanghai, 200433, China.
Abstract:
Soil heavy metal (HM) pollution is a significant and widespread environmental issue in China, highlighting the need to quantify influencing factors and identify priority concern areas for effective prevention and management. Based on published literature data of soil HM concentrations from 2000 to 2022, this study investigated the pollution characteristics and spatial distribution of eight soil HMs in China, and identified the hotspot areas of HM pollution and related influencing factors. The main findings were as follows: (1) The average concentrations of all eight HMs all exceeded their respective background values, with Cd (Igeo = 1.41) and Hg (Igeo = 0.85) showing the most serious pollution. (2) The Random forest-SHapley Additive exPlanations (RF-SHAP) model revealed that transportation and agriculture activities dominantly contribute to soil HM accumulation in China. (3) Bivariate local indicators of spatial association (LISA) based on Moran's I identified industry and transportation activities as primary drivers of HM pollution in the Yangtze River Delta and Pearl River Delta, whereas a combination of agriculture and industry activities was the main cause of pollution in Central China. This study offers valuable insights for the control and management of soil HM pollution and provides a critical reference for shaping comprehensive policies aimed at addressing HM pollution on a regional or national scale.
