Background. Air pollution is a leading cause of mortality and morbidity worldwide. Detailed global assessment of the health effects of short- and long-term exposure to air pollutants requires comprehensive spatio-temporal information on air pollution concentrations over several decades. Currently, empirical measurement data are available for limited time periods only and for a small fraction of the world population. Objectives. To generate, evaluate and analyze a homogeneous multidecadal model hindcast of air quality providing global hourly concentrations for a range of atmospheric pollutants. Methods. We applied the System for Integrated modeLling of Atmospheric coMposition (SILAM) to assess the AQ over the globe in 1980-2015. Results. We produced and evaluated homogeneous hourly time series for O3, NO2, PM2.5, PM10, SO2, CO, NO, NH3 (totally, 86 gaseous and aerosol species) for the period 1980-2015. Discussion. Despite the coarse resolution of the hindcast, inevitable for long-term global simulations, the model skills for several important pollutants, such as PM2.5 and O3, were good, reflecting the comparatively large spatial and long temporal scales of their production from primary pollutants and well-balanced model setup. The trend analysis highlighted the turning points in the pollutant concentration time series. The PM composition reflected the AQ trends in different parts of the world. The computed composition and trends qualitatively agreed with (few) available PM-resolving observations.

2021