We introduce a new method to characterize the influence of atmospheric circulation on surface PM2.5
concentrations. Applying our statistical model to climate projections, we find a strong influence of 2000–2050 climate change on PM2.5
air quality in the United States. We find that current atmospheric chemistry models may underestimate the strong positive sensitivity of PM2.5
to temperature in the eastern United States in summer, and so may underestimate PM2.5
changes in a warmer climate.