High-precision long time series simulation of toxic metallic and metalloid composition of PM2.5 in the Beijing-Tianjin-Hebei Region and surrounding areas of China: 0.01°× 0.01° spatial resolution

Metals and metalloids in PM_2.5 are harmful to human health. However, long-term monitoring data are scarce due to high costs and logistical challenges. This study addresses the scientific problem of the scarcity of continuous long-term spatiotemporal PM_2.5 metals and metalloids data. This study integrated multi-source data to eatablish the random forest models of daily concentrations of 10 metals and metalloids (Al, As, Cd, Cr, Mn, Ni, Pb, Sb, Se, Tl), with high spatial resolution of 0.01°× 0.01°. The test-R² of the models ranged 0.87-0.91. We predicted metals and metalloids in PM_2.5 across the Beijing-Tianjin-Hebei region and its surrounding areas from 2015 to 2020. The concentrations of all 10 components showed a declining trend year by year, with particularly in Al, Pb, and Mn. Seasonally, the highest concentrations occurred in winter, while the lowest were observed in summer. Spatially, higher concentrations were found in Shanxi, southern Hebei, northern Henan, and western Shandong. The high spatialtemporal resolution models constructed and the dataset produced in this study could provide important reference for source control of metal pollution and further study on health effects. SYNOPSIS: This study established high spatiotemporal resolution random forest models and predicted long time series data of 10 metals and metalloids in PM_2.5, thereby solving the scientific problem of limited of big data for monitoring metallic and metalloid components in PM_2.5.