Nitrogen isotopes suggest agricultural and non-agricultural sources contribute equally to NH3 and NH4+ in urban Beijing during December 2018

Agricultural and non-agricultural sources emission contribute to atmospheric ammonia (NH3) and particulate ammonium (NH4+). However, our understanding on the sources of NH3 and NH4+ in PM2.5 (particles smaller than 2.5 μm) during the winter period in the urban atmosphere is limited. Here, we measured the concentrations and stable nitrogen isotopic composition (δ15N) of NH3 and NH4+ in parallel during December 2018 in urban Beijing to assess the non-agricultural and agricultural sources contributions to NH3 and NH4+ in ambient air based on the Chemical Transport Model (CTM), a Bayesian isotope mixing model (SIMMR), and the δ15N signatures that we developed. Our study found weekly NH4+ and NH3 concentrations were on average 2.5 ± 1.4 μg m-3 and 3.8 ± 1.7 μg m-3, respectively during December 2018. Weekly concentration weighted δ15N(NH4+) values ranged from 4.5‰ to 13.7‰ with an average value of 8.2 ± 3.9‰ during December 2018. After accounting for nitrogen isotopic fractionation from NH3 gas to NH4+ conversion, initial δ15N(NH3) values ranged from -22.5‰ to -12.8‰ with an average value of -17.4 ± 3.5‰. Further, weekly measured δ15N(NH3) values ranged from -22.2‰ to -10.2‰ (after correction) with an average value of -15.6 ± 5.3‰ during December 2018. Results from two different isotope-based method showed non-agricultural sources contributed 31.2%-63.1%, with an average value of 47.5 ± 14.6%, to NH4+ and 32.3%-71.2%, with an average of 53.4 ± 16.1%, to ambient NH3 during December 2018 in Beijing. Results from CMAQ-ISAM suggest non-agricultural sources contributed on average 66.2 ± 1.9% to ambient NH4+ and 66.4 ± 1.9% to ambient NH3 during December 2018. Results from this study suggest that agricultural and non-agricultural sources nearly equally contributed to NH3 and NH4+ in urban Beijing during December 2018 with an uncertainty range of 13%-19% between isotope-based methods and CTM method.