Analysis of large-scale multi-annual concentration and exposure simulations for the northern hemisphere

Atmospheric air pollution exerts well-documented impacts on human health, contributing to respiratory and cardiovascular disease and premature mortality at the global scale. In this study, a new module is integrated with the CHIMERE chemistry-transport model to calculate population exposure, exceedances of WHO air quality guidelines, and EPA Air Quality Index values at any given regional scale, and applied to a continuous decade of simulations (2010–2019) over the Northern Hemisphere at 100 km resolution covering six regulated species and their major PM chemical components. Model performance is evaluated against more than 7000 surface monitoring stations, supplemented by aircraft and ozonesonde measurements, demonstrating satisfactory skill for well-mixed species (average correlation 0.68 for O 3 and continental-average correlations of 0.55–0.77 for PM fractions) and expected resolution-related limitations for short-lived primary pollutants (average correlations of around 0.3–0.5 for NO 2 and SO 2 ); the exposure and exceedance metrics reported here are correspondingly presented, for most species, as lower bounds or background exposure values in densely urbanized regions rather than as best estimates of individual exposure. Concentration trend analysis reveals strongly regional patterns: Europe and North America show significant declining SO 2 trends, with Europe also showing declining NO 2 , temporally consistent with the emission control policies implemented over this period, while Asia exhibits a biphasic pattern with a partial reversal after 2013–2014 coinciding with the implementation of Chinese emission control policies. A significant declining trend in ozone and nitrogen dioxide over North Africa is temporally coincident with a significant reduction in regional wind speeds and a weaker, non-significant decline in biomass-burning emissions, though the framework used here does not permit quantitative attribution to a single driver. Wildfire emissions are temporally associated with organic aerosol and black carbon signals across all continents, and most prominently over western North America. Population exposure analysis identifies the Indo-Gangetic Plain as the most intense and consistently worsening PM 2.5 exposure hotspot within the simulated domain. Averaged across Asian grid cells, the WHO PM 2.5 daily guideline (15 μ g m −3 ) is exceeded on approximately 64% of days. A principal contribution of the study is the simultaneous comparison of population-weighted exposure, WHO exceedance frequencies, and EPA AQI distributions computed from the same underlying simulation across six regulated species. This multi-metric comparison shows that AQI distributions remain broadly stable despite meaningful concentration improvements, suggesting that categorical metrics can mask incremental exposure changes at the decadal scale and that continuous exposure trend analysis can provide additional information that is not captured by AQI-type frameworks. The hemispheric decade-long simulation further allows consistent cross-continental confirmation of previously documented regional patterns.