dc.description.abstract | Mercury (Hg) is a toxic metal of global concern due to its long-term persistence in the atmosphere and bioaccumulation in organisms. Various factors could contribute to the difference in ambient Hg concentration among sampling sites, in which meteorological factors could play a predominant role. Previous studies of Lulin Atmospheric Background Station (LABS) mainly focused on the qualitative analysis between atmospheric mercury and meteorological parameters, but lacked quantitative analysis of individual contribution of each meteorological parameter. This study analyzes the ambient Hg concentration and meteorological data at LABS in 2016-2019, and applied it to a generalized additive model (GAM) to evaluate and quantify the perspective contribution of certain meteorological parameters and air pollutants to the variation in ambient Hg concentration.
From 2016 to 2019, mean GEM, GOM, and PBM concentration at LABS were 1.54 ng/m3, 18.4 and 18.2 pg/m3, respectively. Hg diurnal cycle was mainly driven by the mountain-valley breeze which facilitated air mass vertical movement. For seasonal variation, high GEM values in spring could be attributed to the regional transport of biomass burning pollutants from Southeast Asia while high GOM and PBM concentrations in winter were related to strong subsidence events and low rainfall. Air mass origin and transport path difference were the major factors determining the relationship between GEM and temperature, while gas-particle partitioning plays an important role in determining the distribution of GOM and PBM concentrations. Relative humidity (RH) was chosen as an index to represent air mass vertical movement, in which RH showed a correlation with GEM and anti-correlations with GOM/PBM. When RH < 20%, high PBM subsidence events were observed while high GOM subsidence events occurred at 20% < RH < 40%, suggesting the distinct distribution of different Hg species over different atmospheric layers. Besides, strong solar radiation could enhance the photochemical reaction, leading to the positive trend with daytime GOM and PBM concentration.
With the input of aforementioned meteorological parameters, the GAM model can explain 80.2% of the variance of GEM concentration in 2018. However, the fitting result in spring is worse than other seasons due to a rougher time scale of the month parameter and the difference in the composition of pollutants from different emission sources. Among all considered parameters, carbon monoxide has the highest relative contribution, accounting for 47.8% of the total variance of GEM concentration, indicated anthropogenic emission as the main factor in determining/that primarily determined GEM concentration at LABS. Relative humidity has the highest relative contribution among all meteorological factors, representing the dominant contribution of vertical movement of air masses. The Month parameter was used to classify the air mass origin difference caused by seasonal monsoon shifts. Solar radiation exerts a strong positive correlation on GEM, but could weaken under extreme conditions when in-situ oxidation of GEM into GOM is favored. A Negative correlation between GEM and pressure was mainly affected by regional monsoon shifts. Different time-scale mechanisms could weaken the relative contribution of carbon dioxide and temperature on GEM variation. However, by using the interaction between carbon dioxide and solar radiation to investigate GEM concentration difference under different atmospheric conditions, it can be inferred that the absorption process of GEM by vegetation was not the main factor in determining GEM concentration in daytime. | en_US |