Strategies for the coordinated control of particulate matter and carbon dioxide under multiple combined pollution conditions

Air pollutants represented by fine particulate matter (PM2.5) and the greenhouse effect caused by carbon dioxide (CO2), are both urgent threats to public health. Tackling the synergistic reduction of PM2.5 and CO2 is critical to achieving improvements in clean air worldwide. A persistent issue is the identification of their common sources and integrated impacts under different environmental conditions. In this study, we investigated the characteristics of the pollution types captured by combined analysis through a comprehensive observational dataset for 2017-2020, and applied machine learning algorithms to quantify the effects of drivers on air pollutants and CO2 formation. More importantly, detailed conclusions were drawn for the joint control of PM2.5-CO2 in multiple pollution types by using ensemble traceability technique. We demonstrated that reducing coal combustion emissions was an effective measure to maximize the benefits of PM2.5-CO2 in weather with low CO2 levels and no PM2.5 pollution. Correspondingly, on days with severe PM2.5 episodes, prioritizing control of vehicle emissions can simultaneously mitigate PM2.5 and CO2. Similar conclusions were found at high CO2 levels, accompanied by a more extensive role of vehicle emissions. Furthermore, a comparison of the differences in source impacts between PM2.5-CO2 and individual species suggests that focusing only on the sources that contribute significantly to one species may result in an underestimation or overestimation of PM2.5-CO2 source impacts. One such implication, as evidenced by our findings, is that synergistic controlling common sources of pollutants should be efficient. Thereby, common source management targeting PM2.5-CO2 under multiple pollution types is a more workable solution to alleviate environmental pollution.

Automated summary

This study investigates the characteristics of pollution types and quantifies the effects of drivers on air pollutants and CO2 formation through comprehensive observational data and machine learning algorithms. It suggests that controlling common sources of PM2.5-CO2 under multiple pollution types is an efficient solution to alleviate environmental pollution.