Particulate matter (PM2.5) and ozone (O3)
are typical air pollutants that require effective simultaneous control.
Current PM2.5 and O3 co-mitigation strategies
are primarily focused on the emission optimization of nitrogen oxides
(NO
x
) and volatile organic compounds (VOCs);
however, fundamental mechanisms underlying these co-mitigation efforts
remain largely unanswered. Here, utilizing an atmospheric chemistry
model combined with machine learning technique, we extend the scope
of coordinated control efforts beyond just precursors by identifying
and controlling common sources of PM2.5 and O3, which are fossil fuel combustion sources including coal combustion,
industrial emission, and vehicle exhausts. We conducted further sensitivity
simulations and found that reducing fossil fuel combustion emissions
significantly lowers PM2.5 levels but has trade-off effects
on O3 control, which is attributable to the suppressed
aerosol sink of hydroperoxyl radicals and the enhanced atmospheric
oxidizing capacity. Furthermore, this unexpected enhancement of O3 can be offset by synergistically controlling certain other
specific VOCs sources. For example, during days with high levels of
both PM2.5 and O3 pollution, a 40% reduction
of VOCs sources at the corresponding level of coal combustion reduction
is predicted to lead to an ∼13 ppb (∼14%) reduction
in O3 concentration. Our results provide the scientific
evidence and theoretical framework for mitigating PM2.5 and O3 pollution through joint control of fossil fuel
combustion sources and specific VOCs sources.