대기 수은은 가스상으로 존재 가능한 유일한 중금속으로 인체 내 신경계통에 악영향을 미친다. 대기로 배출된 수은은 침적 후 생태계에 영향을 미치며, 특히 수체로 유입된 수은은 혐기성 박테리아에 의해 독성이 높은 메틸수은으로 변환하여 먹이사슬에 따라 고농도로 농축된다. 일반적으로 인간은 어류를 섭취함으로써 수은에 노출된다. 대기 수은은 대부분 무기상태로 존재하며 산화상태에 따라 크게 3가지, 즉 GEM (gaseous element mercury), GOM(gaseous oxidized mercury) 및 PBM(particulate-bound mercury)으로 구분된다. GOM 및 PBM의 농도 및 건식침적속도는 시공간별 변동성이 매우 크기 때문에, 실시간으로 농도를 파악하여 지역적 특성에 따른 규제를 실시하는 것이 중요하다. 국내 연구에서는 수동적인 방법으로 12시간 또는 24시간 간격의 대기 수은 종의 농도를 측정한 바 있으나, 시간적 변이가 큰 대기 수은 종에 대해서는 낮은 시간해상도에 따른 한계점이 존재한다. 따라서 이번 연구의 목적은 현장측정 장비(Tekran speciation system)를 이용하여 시간적 변이가 큰 대기 수은 종의 특성을 파악하고, 이를 바탕으로 수은의 위해도와 직결되는 대기 수은의 건식침적 flux를 추정하는 것이다. 본 연구에서 GEM의 농도는 2019년 6월부터 2020년 12월까지 측정되었으며, GOM 및 PBM의 농도는 2019년 11월부터 2020년 12월까지 측정되었다. GEM은 Tekran 2537X 기기에 의해 5분 간격으로 측정되었으며 GOM 및 PBM은 각각 Model 1130 및 Model 1135에 각각 2시간 채취 후 1시간동안 분석되어 3시간 간격으로 측정되었다. 측정기간 동안의 GEM, PBM, GOM의 평균 농도는 각각 2.5 ± 0.8 ng m-3, 13.1 ± 9.0 pg m-3, 6.2 ± 6.2 pg m-3으로 나타났다. GEM은 대기희석작용이 감소함에 따라 농도가 증가하여 늦은 오후부터 이른 아침까지 높은 농도를 나타내었고, GOM은 낮은 상대습도와 높은 오존의 농도를 보이는 오후시간대에 고농도를 나타내었다. 측정기간 중 상위 10%의 농도가 12시간 이상 지속된 경우를 고농도이벤트로 규정하였으며, 이 경우 대부분 남서기류에서 고농도의 대기 수은이 유입된 것을 확인하였다. 역궤적 기반 수용모델인 potential source contribution function (PSCF) 결과 중국 랴오닝성 부근이 잠재적 배출원으로 파악되었다. 대기 수은 중 GOM 및 PBM은 강우에 의해 유의하게 농도가 감소되는 것을 확인하였으며 강우강도 및 강우량이 증가할수록 농도가 크게 감소하는 것으로 나타났다. 본 연구에서는 3층 저항모델을 사용하여 계산한 건식침적속도와 Tekran speciation system으로 측정한 수은 농도를 이용하여 GOM과 PBM의 토지이용도(land use category: LUC)별 건식침적 flux를 추정하였다. 연구기간동안 대기 농도는 GOM이 PBM보다 낮게 측정되었지만 산정된 건식침적속도는 GOM이 PBM에 비해 높게 계산되어, 최종적으로 추정된 건식침적 flux는 GOM이 PBM에 비해 높게 나타났다. LUC에 따라 추정된 GOM의 건식침적 flux는 도시(424.1 ± 250.0 pg m-2 h-1) > 물표면(370.2 ± 235.1 pg m-2 h-1) > 혼합림(109.5 ± 84.0 pg m-2 h-1) > 농지(99.6 ± 80.3 pg m-2 h-1) 순으로 파악되었으며, 도시 및 물표면에서는 식생에서의 저항이 존재하지 않아 건식침적속도가 높게 산정되었다. PBM의 건식침적 flux는 LUC 및 입경에 따라 다른 특징을 보였다. PBM은 미세입자(PM2.5)에서 높은 농도를 보였으나 건식침적속도는 조대입자(PM2.5-14)에서 높게 산정되었다. 최종적으로 조대입자와 미세입자의 건식침적 flux는 유사한 값을 나타내어, 미세입자 뿐만 아니라 조대 입자 범위에서의 PBM 농도 측정도 중요하다는 것을 시사한다. LUC에 따라 추정된 PBM의 건식 침적 flux는 물표면(83.7 ± 74.6 pg m-2 h-1) > 농지(61.5 ± 55.8 pg m-2 h-1) > 혼합림(57.9 ± 50.2 pg m-2 h-1) > 도시(54.9 ± 49.1 pg m-2 h-1) 순으로 추정되었다. 춘천시 전체 면적에 침적되는 GOM과 PBM의 양은 1,667.3 g yr-1 으로 나타났으며 춘천시 토지는 대부분 산림으로 이루어져 있어 혼합림에 침적되는 수은의 양이 가장 많은 것으로 산정되었다. 이번 연구를 통해 산정된 GOM과 PBM의 건식침적속도는 향후 다양한 지역에서 측정된 대기 중 GOM 및 PBM의 농도와 결합하여 건식침적량의 공간분포를 제시할 수 있으며, 더 나아가 지역 간 수은 위해도 평가 및 대기와 수체간 상관성분석에 대한 연구로 진행될 수 있을 것으로 기대된다.
Atmospheric mercury (Hg) is the only heavy metal that can exist in gaseous form, and it adversely affects the nervous system in the human and the wildlife animals. Hg is emitted to atmosphere from various sources mostly as inorganic forms, but once deposited it is converted to the toxic form, MeHg, by anaerobic bacteria in aquatic ecosystems. MeHg readily bioaccumulates through the food web, and its main exposure pathway for human is consumption of seafood. Atmospheric mercury mainly exists as three forms, including gaseous element mercury (GEM), gaseous oxidized mercury (GOM), and particulate-bound mercury (PBM). GEM is relatively inert and can transport in long-range; however, GOM and PBM have relatively short atmospheric residence time and show the large temporal and spatial variation. Therefore, it is important to measure the Hg concentrations with fine temporal resolution in order to identify the characteristics and possible sources (and/or pathways) of Hg species. In South Korea, atmospheric Hg concentrations were measured at 12 or 24 hour intervals in most studies, having poor temporal resolution; therefore, understanding the fate and transport of GOM and PBM is particularly limited.. In this study, three Hg species were measured using Tekran speciation system, which is in-situ measurement device with high temporal resolution. Samples were collected and analyzed from June 2019 to December 2020 for GEM and from November 2019 to December 2020 for GOM and PBM. GEM was measured at 5-minute intervals by Tekran 2537X instrument, and GOM and PBM were collected for 2 hours on Model 1130 and Model 1135, respectively, and then analyzed for 1 hour, resulting in reporting concentrations at 3-hour intervals. The average concentrations of GEM, PBM, and GOM were analyzed as 2.5 ± 0.8 ng m-3, 13.1 ± 9.0 pg m-3, and 6.2 ± 6.2 pg m-3, respectively, during the study period. GEM concentration increased from late afternoon to early morning as atmospheric dilution decreased. GOM showed high concentration during afternoon with low relative humidity and high ozone (O3) concentration. In this study, high concentration event was defined as when the top 10% concentration lasted more than 12 hours, and most of air parcles were introduced from the southwest. PSCF results for GEM, GOM, and PBM consistently identified Liaoning Province in China as a potential source area. was In addition, it was shown that GOM and PBM concentration statistically decreased via wet-deposition, and their concentration reduction was correlated with precipitation depth and rainfall intensity. In this study, dry deposition fluxes of GOM and PBM were calculated based on measured atmospheric concentration. Dry deposition velocities (Vd) were estimated using three-layer resistance model. GOM showed higher Vd than PBM, resulting in higher deposition flux than PBM, although atmospheric concentrations of GOM were generally lower than those of PBM. Dry deposition fluxes of GOM were 424.1 ± 250.0 pg m-2 h-1(urban), 370.2 ± 235.1 pg m-2 h-1(water), 109.5 ± 84.0 pg m-2 h-1 (forest), and 99.6 ± 80.3 pg m-2 h-1(farm field and grasses). PBM showed higher concentration in fine mode than in coarse mode while Vd was higher in coarse PBM than fine PBM, causing similar dry deposition flux between PBM in fine mode and PBM in coarse mode. This result indicates that the concentration of PBM in coarse mode should be measured to estimate the total dry deposition flux of PBM. Dry deposition flux of PBM was the highest in water surface (83.7 ± 74.6 pg m-2 h-1), followed by farm field (61.5 ± 55.8 pg m-2 h-1), forest (57.9 ± 50.2 pg m-2 h-1), and urban (54.9 ± 49.1 pg m-2 h-1) in this study. Total deposition of GOM and PBM into city of Chuncheon was estimated to be1,667.3 g yr-1. The dry deposition velocities of GOM and BPM calculated in this study can be combined with the atmospheric Hg concentrations in various regions to suggest the spatial distribution of dry deposition flux.
목 차I. 서론 11. 수은의 물리?화학적 특성 12. 수은의 주요 배출원 23. 수은의 변형 및 환경매체 간 이동 34. 수은의 연구현황 44.1. 국외 연구현황 44.2. 국내 연구현황 55. 연구목적 및 필요성 7II. 연구방법 91. 시료채취 장소 및 채취기간 92. 시료 채취 방법 103. 건식 침적 flux 추정 123.1. GOM의 건식침적속도 산정 123.2. PBM의 입경별 농도 및 건식침적속도 산정 143.3. 춘천시에 침적된 수은의 질량 산정 164. HYSPLIT 모델을 활용한 배출원 분석 174.1. 역궤적 분석(back trajectory analysis) 174.2. PBL(planetary boundary layer) 174.3. PSCF 분석(potential source contribution function) 175. 대기오염물질 자료 수집 및 통계분석 196. QA/QC 20III. 결과 및 토의 231. 대기 중 수은의 시간에 따른 농도특성 231.1. 연구기간동안의 대기 수은 농도 231.2. 월평균 및 계절적 변화 281.3. 대기 수은 종의 일변화 332. 특정 이벤트에 따른 대기 수은농도의 변화 392.1. 강수에 의한 수은 농도변화 392.2. 대기 수은 종의 고농도 이벤트 432.3. GEM 432.4. PBM 442.5. GOM 442.5. PSCF 503. 대기 중 수은의 건식침적 flux 파악 533.1. GOM 및 PBM의 건식침적속도 573.2. GOM 및 PBM의 건식침적 flux 633.3. 측정기간동안 침적된 수은의 질량(amount) 73IV. 결론 76□ 참고문헌 78□ Abstract 91List of TableTable 1. Mercury concentrations reported in previous studies 4Table 2. Air pollutant emission facility in Chuncheon (reported in as of 2018) 9Table 3. Size-segregated PBM concentration percentage from 0 to 14 μm in previous study (Kim, 2017) 15Table 4. Land use categories classified in this study. 16Table 5. Standard Operating Procedures in various countries for atmospheric mercury concentration measurement using Tekran speciation system 21Table 6. Concentrations of atmospheric mercury species in South Korea reported in other studies 25Table 7. Averaged seasonal concentrations of atmospheric mercury species 29Table 8. Averaged GEM concentration for each cluster during June and July in 2019 and in 2020 32Table 9. Factor analysis results 38Table 10. Averaged concentrations of the three Hg species and representive air pollutants for each cluster of GEM, GOM, and PBM high concentration event 49Table 11. Pearson correlation coefficients between dry deposition flux, concentration, and deposition velocity for PBM by land use category (LUC) 54Table 12. Pearson correlation coefficient between dry deposition flux, concentration, and deposition velocity for GOM by LUC 54Table 13. Dry deposition flux of GOM and PBM in previous studies 56Table 14. Dry deposition velocity by particle size in the urban during the measurement period 59Table 15. Dry deposition velocity by particle size in mixed farm during the measurement period 60Table 16. Dry deposition velocity by particle size on the water surface during the measurement period 61Table 17. Dry deposition velocity by grain size in mixed forest during the measurement period 62Table 18. Monthly averaged and seasonally averaged dry deposition velocity and dry deposition flux for GOM 65Table 19. PBM Dry deposition flux by particle size in urban during the measurement period 69Table 20. PBM dry deposition flux by particle size in mixed farm during the measurement period 70Table 21. PBM Dry deposition flux by particle size in water surface during the measurement period 71Table 22. PBM dry deposition flux by particle size in mixed forest surface during the measurement period 72Table 23. GOM deposited amount for each LUC in Chuncheon, South Korea 74Table 24. PBM deposited amount for each LUC in Chuncheon, South Korea 75List of FiguresFig. 1. Location of the sampling site. 9Fig. 2. Schematic diagram of the Tekran speciation system(source : Laurier et al., 2004). 11Fig. 3. Concentrations of hourly GEM (top panel), bihourly GOM (middle panel), and bihourly PBM (bottom panel) during the whole sampling period. 26Fig. 4. Concentration distribution of atmospheric mercury species during the measurement period(top-GEM, middle-GOM, bottom-PBM). 27Fig. 5. Box-plot of monthly GEM concentration. Solid red line indicates the arithmetic mean. 30Fig. 6. Box-plots of monthly PBM (upper panel) and monthly GOM (lower panel) concentration. 31Fig. 7. Mean back-trajectories of each cluster during June and July in 2019(left) and 2020(right). 32Fig. 8. Diel variations of GEM (top), GOM (middle), and PBM (bottom) for each season. 35Fig. 9. Diel variations of O3 (top), atmospheric relative humidity (middle), and wind speed (bottom) for each season. 36Fig. 10. The variation in atmospheric Hg (GEM, PBM, and GOM) and ventilation coefficient. 37Fig. 11. Average decrease rate according to rainfall event. 41Fig. 12. Concentration reduction rate as a function of rainfall intensity. 41Fig. 13. Concentration reduction rate as a function of precipitation depth. 42Fig. 14. Concentration ratio before and after rain event as a function of precipitation depth. 42Fig. 15. Back-trajectory cluster analysis for GEM high concentration event. 46Fig. 16. Back-trajectory cluster analysis for PBM high concentration event. 47Fig. 17. Back-trajectory cluster analysis for GOM high concentration event. 48Fig. 18. PSCF results for the top 25% GEM (upper) and for the top 25% GOM (lower) concentrations. 51Fig. 19. PSCF result for the top 25% PBM concentration. 52Fig. 20. GOM concentration (top panel), dry deposition velocity (Vd) (the second panel from above), the dry deposition flux calculated using daily averaged concentration and daily averaged Vd (the third panel from above), and the dry deposition flux calculated using hourly concentration and hourly Vd (bottom panel). 55Fig. 21. Dry deposition velocity (line graph) according to LUC for each particle size and size-segregated PBM concentration(bar graph). 58Fig. 22. Dry deposition flux of GOM during the measurement period. 64Fig. 23. Dry deposition flux by PBM size according to LUC. 67Fig. 24. Dry deposition flux of PBM during the measurement period. 68Fig. 25. Contribution rate of each LUC to total deposited amount of GOM (left) and PBM (right) in Chuncheon, South Korea. 73