Trends in PM air pollution in the context of internal climate variability*

Inhalation of particulate matter (PM) air pollution is a leading cause of premature mortality worldwide. There have been large, regionally distinct changes in regional anthropogenic emissions of PM and its precursors over the past few decades. Observational records indeed indicate substantial changes in PM air pollution, and we rely on such records to probe the efficiency of anthropogenic emission controls and advance process understanding of PM. However, observational records are influenced to an unknown extent by climate variability due to strong connections between PM and meteorology through sources and sinks (e.g., transport, wet and dry deposition, precursor emissions from vegetation and fire). Here we use examine historical simulations from two different interactive aerosol configurations of the NASA GISS global climate model, with ten ensemble members each, as well as the fifteen-member ensemble of historical simulations from CESM2 with interactive aerosols. We establish whether the set of simulations captures observed PM trends since the 1990s as well as the observed range of interannual variations, and use this analysis to undercover model strengths and limitations. In parallel, we indicate regions and seasons where the forced trend emerges strongly, or weakly, from meteorological noise.