Distinguishing Emission-Associated Ambient Air PM_2.5 Concentrations and Meteorological Factor-Induced Fluctuations

Although PM_2.5 (particulate matter with aerodynamic diameters less than 2.5 μm) in the air originates from emissions, its concentrations are often affected by confounding meteorological effects. Therefore, direct comparisons of PM_2.5 concentrations made across two periods, which are commonly used by environmental protection administrations to measure the effectiveness of mitigation efforts, can be misleading. Here, we developed a two-step method to distinguish the significance of emissions and meteorological factors and assess the effectiveness of emission mitigation efforts. We modeled ambient PM_2.5 concentrations from 1980 to 2014 based on three conditional scenarios: realistic conditions, fixed emissions, and fixed meteorology. The differences found between the model outputs were analyzed to quantify the relative contributions of emissions and meteorological factors. Emission-related gridded PM_2.5 concentrations excluding the meteorological effects were predicted using multivariate regression models, whereas meteorological confounding effects on PM_2.5 fluctuations were characterized by probabilistic functions. When the regression models and probabilistic functions were combined, fluctuations in the PM_2.5 concentrations induced by emissions and meteorological factors were quantified for all model grid cells and regions. The method was then applied to assess the historical and future trends of PM_2.5 concentrations and potential fluctuations on global, national, and city scales. The proposed method may thus be used to assess the effectiveness of mitigation actions.