Concentrated fireworks display-induced changes in aerosol vertical characteristics and atmospheric pollutant emissions

Fireworks affect the atmospheric environment and human health due to the rapid release of high levels of air pollutants. It is crucial to clarify the vertical characteristics of air pollutants and quantify the emission quantity for accurate control and health assessment. With the use of an unmanned aerial vehicle and a ground-based LiDAR system, we investigated the vertical profiles of particulate matter (PM) under the policy of concentrated fireworks detonation in a megacity (i.e., Changchun) in Northeast China and developed emission factors for fireworks displays. The results showed that the PM_2.5, PM_10, SO₂ concentrations within a height of 200 m in the core area of fireworks display significantly increased by 143.8 %, 63.7 % and 174.3 %, respectively, on the day of the Lantern Festival relative to the background concentrations. There was a decreasing trend in the PM_2.5 concentration with increasing height, with a maximum vertical decrease of 64 μg/m³ per 100 m. Despite a general decline in the concentration, the vertical distribution of polluted air masses exhibited heterogeneity, occasionally exhibiting a bimodal distribution pattern. Moreover, the correlation coefficients between the aerosol optical depth (AOD) at different altitudes (0.2–0.4, 0.4–0.7, and 0.7–1.0 km) and the ground-level PM concentration were 0.29, −0.35, and −0.35, respectively. This phenomenon indicates a reduced influence of ground-level PM emission on the AOD at altitudes from 0.4 to 1.0 km and the increased potential for aerosol particles to escape upwards, particularly in the troposphere. Furthermore, an approach combining real-time stereoscopic monitoring of PM and downwind wind speed was developed to calculate the estimated firework emission factors. The emission factors recorded during the Lantern Festival, specifically between 20:00 and 24:00, were 128.6 ± 28.3 mg/m² per 10 min and 92.7 ± 23.2 mg/m² per 10 min for PM_2.5 and PM₁₀, respectively, including an uncertainty of 9 %–14 % generated by the sensors. The proposed approach and corresponding emission factors could be used to easily quantify air pollutant emissions originating from fireworks displays and to conduct further numerical simulations for air pollution regulation and health assessment.