목차

목 차
Ⅰ. 서론 1
1. PM2.5 정의 및 구성성분 1
2. 바이오매스 연소 3
3. 연구배경 5
4. 연구목적 8
Ⅱ. 실험방법 9
1. 시료 채취 9
2. PM2.5 질량 농도 11
3. 수용성 무기이온 11
4. 유기탄소(OC)와 무기탄소(EC) 13
5. Levoglucosan, Mannosan and Galactosan 15
6. 기상 및 기타 분석 자료 21
7. 정도 관리(Quarlity Assurance and Quality Control, QA/QC) 22
Ⅲ. 결과 26
1. PM2.5 질량농도 26
2. 수용성 이온성분 29
3. 탄소성분 34
1) 유기탄소(OC)와 무기탄소(EC) 34
2) Levoglucosan, Mannosan and Galactosan 41
3) 바이오매스 연소 영향 확인 46
4) 1차 유기탄소 및 2차 유기탄소 추정 53
5) 장거리 이동 바이오매스 연소 영향 59
4. 고농도 이벤트(High Concentration Episodes, HCE) 61
5. 주성분 분석(principal component analysis, PCA) 69
Ⅳ. 결론 72
□ 참고문헌 74
□ Abstract 91
표 목 차
Table 1. Information of levoglucosan, mannosan and galactosan 4
Table 2. Sampling information 10
Table 3. Analysis conditions of ion chromatography 12
Table 4. Experimental conditions for NISOH 5040 method 14
Table 5. GC/MS operating conditions 17
Table 6. Information of target polar compounds 18
Table 7. Summarized QA/QC results for PM2.5 component 23
Table 8. Result of calibration parameter (r2) and MDL (ng/m3) oflevoglucosan, mannosan and galactosan 24
Table 9. Results of recovery (%) of methyl β-D-xyropyranoside, a internal standard of levoglucosan, mannosan and galactosan 24
Table 10. Results of RSD (relative standard deviation, %) of levoglucosan, mannosan and galactosan 25
Table 11. Results of ESC (external standard check, %) of methyl β-D-xyropyranoside, levoglucosan, mannosan and galactosan 25
Table 12. Seasonally averaged PM2.5 mass concentration 27
Table 13. Seasonal average concentrations of ionic components in PM2.5 30
Table 14. Seasonal average concentrations of OC and EC 36
Table 15. Seasonal average concentration of OC and EC fraction 37
Table 16. Seasonal contribution of each OC fraction to total OC (%) 38
Table 17. Seasonal average concentrations of levoglucosan, mannosan and galactosan 42
Table 18. Seasonally estimated concentrations of POC and SOC 58
Table 19. Pearson correlation coefficients between PM2.5 components inHCE samples. Numbers in bold with two stars and with onestar indicate that the correlations were siginificant at the 0.01and 0.05 levels, respectly. 65
Table 20. Pearson correlation coefficients between PM2.5 components innon-HCE samples. Numbers in bold with two stars and withone star indicate that the correlations were siginificant at the0.01 and 0.05 levels, respectly. 66
Table 21. Factor loadings in PM2.5 components with representativegaseous polluants and meteorological parameters in principalcomponent analysis (PCA) 70
Table 22. Pearson correlation coefficients between OCbb and PM2.5component. Numbers in bold with two stars and with one starindicate that the correlations were siginificant at the 0.01 and0.05 levels, respectly. 71
그 림 목 차
Fig. 1. Annual emission rates of PM2.5 according to Clean Air Policy Support System (CAPSS) and annual average concentration of PM2.5 in major cities in Korea. 6
Fig. 2. Contribution of each emission sector to total PM2.5 emission in Chuncheon (CAPSS, 2017). 7
Fig. 3. Mass spectra of (a) methyl β-D-xyropyranoside, (b) levoglucosan, (c) mannosan and (d) galactosan. 19
Fig. 4. Ion chromatogram in SIM mode (m/z : 73, 204, 217) for (a)standard, (b) blank, (c) October 14, 2018 and (d) January 14,2019. 17
Fig. 5. Daily concentration of PM2.5 during the whole sampling period. 28
Fig. 6. Contribution of each ionic component to PM2.5 mass (%) duringthe whole sampling period. 31
Fig. 7. Contributions of each ionic component to PM2.5 mass (%) for eachseason. 32
Fig. 8. Correlation between 2[SO42-]+[NO3-] and [NH4+]. 33
Fig. 9. Correlation between EC and OC concentration. 39
Fig. 10. Relationship between OC and EC for each season. 40
Fig. 11. Temporal variation of levoglucosan, mannosan and galactosanduring the whole sampling period. 43
Fig. 12. Correlations between levoglucosan and galactosan (left panel) and between levoglucosan and mannosan (right panel). Blue, red, yellow, and green areas indicate the hardwood burning, softwood burning, grass burning, and crop residue burning, respectively. 44
Fig. 13. The LG/GA ratio (upper panel) and the LG/MN ratio (lower panel) for the whole sampling period. The blue, red, yellow, and green areas indicate the burning of hardwood, softwood, grass, and crop residue, respectively. 45
Fig. 14. Correlation between levoglucosan and K+ for (a) whole samplesand (b) without January 14, 2019 sample. 48
Fig. 15. Correlations between levoglucosan and K+ for each season. 49
Fig. 16. Correlation between levoglucosan and OC (left panel) andbetween levoglucosan and EC (right panel). 50
Fig. 17. Correlation between levoglucosan and OC (upper panel) and between levoglucosan and EC (lower panel). The yellow and purple symbols indicate the samples obtained during spring and summer and during fall and winter, respectively.. 51
Fig. 18. The percentage of levoglucosan in OC for each season. 52
Fig. 19. Correlations between OC and EC. The regression equation wasobtained using the lowest 10% of OC/EC ratio (red symbol) toestimate the primary organic carbon (POC). 56
Fig. 20. Concentrations (bar graphs) and contribution (symbol graphs) ofOC (upper panel) and EC emitted from biomass burning andvehicle. Concentration and fraction of SOC to OC were alsoindicated in upper panel. 57
Fig. 21. Air parcel trajectories arriving at the sampling site whenhighlevoglucosan and low NOx concentrations appeared (red symbol :fire spot). 60
Fig. 22. Comparison between the PM2.5 components of the HCE andnon-HCE samples. The red symbols indicate the increment ratioof each PM2.5 component during the HCE. 63
Fig. 23. Fraction of each component in the PM2.5 mass (left) and thefraction of each sugar component in the OC mass (right) forHCE and non-HCE samples. 64
Fig. 24. Backward trajectories for high concentration episodes (red symbol : fire spot). 67
Fig. 25. Comparison of the PM2.5 components collected on January 14,2019 with the average concentrations of constituents of PM2.5collected over the entire sampling period.. 68