Abstract
Based on hourly observation data from the aethalometer and GRIMM180 environment particle monitor as well as the simultaneous data of visibility (V), relative humidity (RH) and nitrogen dioxide (NO2) from October to December in 2017 in Chengdu, the corresponding time series of aerosol extinction coefficient per unit of mass is retrieved. The generalized additive models (GAMs) are adopted to analyze the non-stationarity of the time series of aerosol extinction coefficient per unit of mass and to explore the responses of the aerosol extinction coefficient per unit of mass to the aerosol component structure factors (ρBC/ρPM10, ρPM1/ρPM2.5, ρPM1~2.5/ρPM2.5 and ρPM2.5/ρPM10; ρ represents particle mass concentration) and RH. The results show that through the comparative analysis of stationary and non-stationary models, the time series of aerosol extinction coefficient per unit of mass in autumn and winter in Chengdu is non-stationary. In addition, the RH and aerosol component structure factors are all significant nonlinear covariates that affect the non-stationarity of the aerosol extinction coefficient per unit of mass. According to the influence of covariates, the sequence is as follows: RH > ρBC/ρPM10 > ρPM2.5/ρPM10 > ρPM1/ρPM2.5. At PM2.5 pollution concentration (ρPM2.5 > 75 μg m−3), according to the influence of covariates, the sequence is as follows: RH > ρPM1~2.5/ρPM2.5 > ρBC/ρPM10 > ρPM2.5/ρPM10. Moreover, the interaction between RH and aerosol component structure factors significantly affects the aerosol extinction coefficient per unit of mass. The condition of high RH, high ρPM2.5/ρPM10, high ρPM1/ρPM2.5 and low ρBC/ρPM10 has a synergistic amplification effect on the increase of the aerosol extinction coefficient per unit of mass. At PM2.5 pollution concentration, the synergistic effect of high RH, high ρPM2.5/ρPM10, high ρPM1~2.5/ρPM2.5 and low ρBC/ρPM10 is beneficial to the increase of the aerosol extinction coefficient per unit of mass.
Funder
the National Key R&D Program of China
Subject
Atmospheric Science,Environmental Science (miscellaneous)
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献