We interpret in situ and satellite observations with a chemical transport model (GEOS-Chem, downscaled to 0.1° × 0.1°) to understand global trends in population-weighted mean chemical composition of fine particulate matter (PM2.5). Trends in observed and simulated population-weighted mean PM2.5 composition over 1989−2013 are highly consistent for PM2.5 (−2.4 vs −2.4%/yr), secondary inorganic aerosols (−4.3 vs −4.1%/yr), organic aerosols (OA, −3.6 vs −3.0%/yr) and black carbon (−4.3 vs −3.9%/yr) over North America, as well as for sulfate (−4.7 vs −5.8%/yr) over Europe. Simulated trends over 1998−2013 also have overlapping 95% confidence intervals with satellitederived
trends in population-weighted mean PM2.5 for 20 of 21 global regions. Over 1989−2013, most (79%) of the simulated increase in global populationweighted
mean PM2.5 of 0.28 μg m−3 yr−1 is explained by significantly (p < 0.05)
increasing OA (0.10 μg m−3 yr−1 ), nitrate (0.05 μg m−3 yr−1), sulfate (0.04 μg
m−3 yr−1 ), and ammonium (0.03 μg m−3 yr−1 ). These four components predominantly drive trends in population-weighted mean PM2.5 over populous regions of South Asia (0.94 μg m−3 yr−1), East Asia (0.66 μg m−3 yr−1), Western Europe (−0.47 μg m−3 yr−1), and North America (−0.32 μg m−3 yr−1). Trends in area-weighted mean and population-weighted mean PM2.5 composition differ significantly.