Summertime Urban Ammonia Emissions May Be Substantially Underestimated in Beijing, China

Ammonia (NH3) is critical to the nitrogen cycle and PM2.5 formation, yet a great deal of uncertainty exists in its urban emission quantifications. Model underestimated NH3 concentrations have been reported for cities, yet few studies have provided an explanation. Here, we explore reasons for severe WRF-Chem model underestimations of NH3 concentrations in Beijing in August 2018, including simulated gas-particle partitioning, meteorology, regional transport, and emissions, using spatially refined (3 km resolution) NH3 emission estimates in the agricultural sector for Beijing Tianjin-Hebei and in the traffic sector for Beijing. We find that simulated NH3 concentrations are significantly lower than ground-based and satellite observations during August in Beijing, while wintertime underestimations are much more moderate. Further analyses and sensitivity experiments show that such discrepancies cannot be attributed to factors other than biases in NH3 emissions. Using site measurements as constraints, we estimate that both agricultural and non-agricultural NH3 emission totals in Beijing shall increase by similar to 5 times to match the observations. Future research should be performed to allocate underestimations to urban fertilizer, power, traffic, or residential sources. Dense and regular urban NH3 observations are necessary to constrain and validate bottom-up inventories and NHx simulation.

Automated summary

This study investigates the reasons for the underestimations of NH3 concentrations in Beijing in August 2018 by the WRF-Chem model, including simulated gas-particle partitioning, meteorology, regional transport, and emissions. The results show that the simulated NH3 concentrations are significantly lower than the ground-based and satellite observations, with more moderate underestimations in winter. Further analyses and sensitivity experiments indicate that these discrepancies cannot be attributed to factors other than biases in NH3 emissions. The study suggests that both agricultural and non-agricultural NH3 emissions in Beijing need to increase by around 5 times to match the observations, and future research should focus on allocating the underestimations to urban fertilizer, power, traffic, or residential sources. Dense and regular urban NH3 observations are necessary to validate the bottom-up inventories and NHx simulation.