An integrated approach for determining the spatial distribution, source apportionment and human health risks of PAHs in soils of rapidly-urbanized areas

The rise of large urban agglomerations has exacerbated pollutant emissions, resulting in regional soil contamination with polycyclic aromatic hydrocarbons (PAHs), which jeopardizes the development of urban agglomerations and affects human health. There is a lack of research in this area in the Beijing-Tianjin-Hebei region, with the existing studies on PAHs in soil in the larger region often neglecting the spatial heterogeneity of the pollution sources and systematic analysis of risk assessment. This study introduces the Distribution-Source-Risk framework, analyzed soil PAH pollutants in the region, and examined PAH sources, distribution patterns, and associated health risks. Random forest modeling was employed to map PAH distribution in the BTH soils. City classification analysis was conducted based on the derived pollution levels and urbanization degree, resulting in four city types: high urbanization and high pollution, high urbanization and low pollution, low urbanization and high pollution, and low urbanization and low pollution. Primary PAH sources include coal-burning (29%), coking (25%), traffic (25%), and biomass-burning (21%), with varying contributions based on city types. The overall order of human health risks was coal-burning ​> ​traffic ​> ​coking ​> ​biomass-burning sources. Finally, differen policies for soil PAH PAH management (such as energy transition and green infrastructure) were elaborated to promote coordinated development of regional urbanization environment. In summary, this research offers a comprehensive approach, linking processes to provide a precise understanding of pollution across different entities (cities, sources, and populations). Our findings reveal distinct pollution patterns across city types and highlight targeted mitigation priorities and provide a systematic, data-driven framework for regional soil PAH management and public health protection. • Established a universal method framework for Distribution-Source-Risk. • Machine learning reveals high levels of pollution in the NE/SW areas. • The sources of pollution vary among different cities. • Traffic and coal-burning sources pose high risks. • Average carcinogenic risk of soil PAHs is close to the threshold 1E-04.