Deriving Non‐Methane Hydrocarbon Emissions and Improving Source Apportionment by the Boundary Layer Mass Balance Technique

Abstract Although high‐resolution emission inventory (EI) has been compiled, there are still large uncertainties. This study employs a boundary layer mass balance (MB) method (also known as the hypothetical box model) to calculate the nonmethane hydrocarbons (NMHCs) emission fluxes based on the ground surface and tower measurements between September and November 2018 in the urban core of the Pearl River Delta (PRD) region, China. The MB method estimated the propane, toluene, n‐butane, m,p‐xylene, and i‐butane as the major species, which are in reasonable agreement with the estimates by mixed layer gradient technique. The physical dispersion term explained over 80% of the emission flux variations for most species while the chemical loss term took up nearly 60% for the highly reactive species (e.g., isoprene and styrene). Higher emission fluxes were identified during daytime than nighttime, well reflecting the daily activities. The flux‐based positive matrix factorization (FLU‐PMF) and traditional concentration‐based apportionment both revealed that vehicular emissions and oil and gas evaporation were the dominant sources of NMHC emissions. FLU‐PMF attributed a higher contribution to biogenic emissions and a reduced contribution to fuel combustion by accounting for chemical loss and atmospheric dilution. Higher NMHC emission flux was found by MB estimates than EI grids near the measurement sites. It was suggested that EI underestimated the emission fluxes due to missing information for the residential fugitive emissions, while industrial contributions were overestimated. The findings provide useful information for policy‐makers and government to better constrain volatile organic compound emissions and improving the source apportionment in urban areas.