Impacts of the COVID-19 lockdown in China on new particle formation and particle number size distribution in three regional background sites in Asian continental outflow

Despite the curtailment of atmospheric condensing precursor gases during the Coronavirus disease 2019 (COVID-19) lockdown (LD) period, unexpected haze events via the formation of new particles and their subsequent growth have been identified. This study investigated the impact of emission reduction during the Chinese LD period on the new particle formation (NPF) frequency and corresponding particle number size distribution (PNSD) at three regional background atmospheric monitoring sites in the western coastal areas of the Korean Peninsula. During this duration, the number concentrations of the nucleation- (<25 nm) and accumulation-mode (>90 nm) particles significantly decreased in Baengryeong (BRY), showing decreases of 34% and 29%, respectively. Unlike BRY, the PNSD in Anmyeon (AMY), which is influenced by nearby industrial emissions, remained nearly unchanged during the LD period, possibly because the reduction in industrial emissions was not significant during the social distancing period enforced by Korea. Bongseong (BOS) showed a similar variation to that of BRY; however, the magnitude of the reduction was weaker because of its higher altitude compared to other sites. The cyclostationary empirical orthogonal function technique was applied to the measured PNSDs at the three sites to objectively classify NPF events. Because mode 1 of cyclostationary loading vectors commonly represented the typical diurnal variation of PNSD during regional NPF events at three sites, mode 1 of the corresponding principal component time series was used for NPF classification. The NPF frequency decreased by 7%, 1%, and 7% in BRY, AMY, and BOS, respectively, despite favorable meteorological conditions, such as increased temperature and insolation during the LD period. The diurnal variation in the sulfuric acid (H(2)SO(4)) proxy implied that the H(2)SO(4) proxy acted as a determining factor for NPF events during the NPF occurrence time (8–12 local hours) in AMY and BOS; however, NPF occurrence in BRY was not connected to the H(2)SO(4) proxy level. This suggests that BRY was more likely to be influenced by the reduction in organic species in the continental upwind regions, while the occurrence of NPF events in AMY and BOS can be suppressed in association with the distinct reduction in inorganic compounds represented by the H(2)SO(4) proxy during the LD period.