Improving PM2.5 simulations in the Metropolitan Area of São Paulo through PMF-based emission inventory adjustments and local road transport emissions

According to the source apportionment receptor model, the road transport sector is the main source of fine particulate matter (PM 2.5 ) concentrations in the city of São Paulo. However, air quality models running with local or global emission inventories underestimate PM 2.5 concentrations. Combining experimental data and receptor model results, this study develops a detailed local emission inventory for the road transport sector that accounts for road dust resuspension and São Paulo’s biofuel composition. Emissions from other anthropogenic sectors were estimated using global datasets and adjusted based on previous source apportionment studies that relied on the positive matrix factorization modeling conducted in the city of São Paulo. This updated emission inventory was then used in the Community Multiscale Air Quality (CMAQ) model over a two-month period (April–May 2018). The meteorological model using objective analysis performed well at two ground-based stations. Although the CMAQ model PM 2.5 predictions performed better with the newer inventory, ozone predictions showed lower modeling performance than with the global inventory. Predicted secondary aerosols and chemical components of PM 2.5 revealed agreements with observations, specifically elemental carbon and sulfates, as well as discrepancies for other components. The enhanced emission inventory improved model accuracy for PM 2.5 , sulfur dioxide, and carbon monoxide by accounting for not only the road transport sector but also other relevant sectors. This inventory will serve as a baseline for a future analysis in a scenario projecting an increase in biofuel use.