Machine learning maps PFAS contamination risk in China's surface water

A machine learning model mapped PFAS exceedance risk across China's surface water, identifying high-risk hotspots in eastern coastal and key inland industrial provinces. An estimated 80-90 million people live in these high-risk areas.

Researchers developed a Geographically Weighted Random Forest (GWR-RF) model to map PFAS exceedance risk in China's surface water, integrating data from over 280,000 potential PFAS sources to overcome sparse monitoring data.

This research provides scientific evidence to support strategic actions for managing PFAS contamination and protecting public health.

This cross-sectional study used a machine learning approach to map the risk of Per- and polyfluoroalkyl substances (PFAS) exceedance in China's surface water. It integrated a comprehensive inventory of 280,000 potential PFAS sources to address sparse monitoring data.