| 摘要: | 每年3月至4月中南半島北部山區會發生大規模生質燃燒，燃燒煙團上升到高海拔大氣層後，隨著盛行西風傳輸至東亞，受影響地理範圍很大；當傳輸煙團與雲層交會，將影響雲滴輻射收支，對氣候變遷影響重大。
本文於2015年3~4月泰國安康山進行密集觀測，PM2.5和PM10質量濃度分別為88.7±36.1 和112.0±39.0 µg m-3，PM2.5占PM10質量濃度的79%， PM2.5最重要化學成分類別為有機碳(OC)，占PM2.5質量濃度37.2%，碳成分中OC3和EC1-OP分別占OC和元素碳(EC)的34.2%與93%，可視為生質燃燒指標物種；OC有高達68.9%為水可溶有機碳(WSOC)，當這些氣膠進入雲層，將改變雲滴化學性質，因而影響雲滴輻射收支；在近生質燃燒污染源區，左旋葡聚糖(Levoglucosan)有突出的占比。本文探討氣膠化學成分比值應用，發現char-EC/soot-EC比OC/EC更能判定生質燃燒影響以及燃燒程度；使用Levoglucosan/Mannosan比值，可推論中南半島燃燒材質為軟木。
相同期間鹿林山觀測PM2.5和PM10質量濃度分別為21.1±9.6 和25.1±10.9 µg m-3，PM2.5占PM10質量濃度的84.2%。nss-SO42-和OC分別占PM2.5質量濃度的14.2和24.0%，碳成分中OC3和EC1-OP分別占OC和元素碳(EC)的36.5%與77.0%。在非生質燃燒期間，PM2.5最重要成分是nss-SO42-，占PM2.5質量濃度的24.8%，OC占PM2.5質量濃度22.8 %，OC仍然是以OC3占比最高(30.2%)，EC則是以EC2占比(68.0%)最高，顯示非生質燃燒期間，重要氣膠成分和生質燃燒期間稍有不同。另外，利用成分比值推估鹿林山在非生質燃燒期間受汽、機車或燃煤影響大，這指出了背景大氣氣膠主要污染源。在其他物種方面，nss-K+、NO3-、單醣無水化合物、二元酸及其鹽類濃度在生質燃燒影響期間相較於非生質燃燒期間濃度凸顯，可視為生質燃燒指標物。
檢視近生質燃燒源與經長程傳輸煙團氣膠成分質量濃度占比，以nss-K+最為穩定，在PM2.5占比都是2%，本文以Modification Factor (MF)判別氣膠成分在長程傳輸後的增益或衰減，發現nss-SO42-在傳輸過程增益最大，顯示傳輸過程中有既成的nss-SO42-氣膠加入或有前驅氣體轉化現象的發生，單醣無水化合物在傳輸過程衰減最為嚴重，表示長時間在大氣停留，單醣無水化合物可能會漸漸消失。
大氣氣膠採樣影響因素很多，本文發現近生質燃燒源採樣過程中， OC1碳成分量測受到石英濾紙吸附揮發性有機氣體的干擾最大，在PM2.5小於77 µg m-3時差異顯著。生質燃燒煙團經傳輸後以一張、兩張或三張石英濾紙採樣，OC1都有顯著性的差異。在水可溶無機離子方面，Cl-揮發影響比例最高，近生質燃燒源和生質燃燒煙團傳輸後，修正後和未修正Cl-濃度分別為3倍及2.2倍。
以PMF解析中南半島污染來源共可得到4類，有72%氣膠受到生質燃燒影響，大致上與逆推軌跡來向分類具有一致性，但人為源污染排放及海鹽影響仍有其影響；以PMF解析鹿林山污染來源共可得到6類，有77.8%氣膠受到生質燃燒影響，確認生質燃燒煙團長程傳輸的現象。
;Every spring, biomass burning (BB) occurs extensively in the mountain area of northern Indochina. The plume produced from BB is uplifted to the elevated atmosphere and transported by prevailing westerly from Indochina to East Asia. As the BB plume distributes spatially, it will affect solar radiation budget when mixing with cloud layers during transport and thus resulting in an effect on climate change in the region.
In this study, atmospheric aerosol was observed intensively at Mt. Doi AnKhang (DAK) in Chiang Mai, Thailand from March to April 2015. During the observation period, PM2.5 and PM10 mass levels were averaged at 88.7±36.1 and 112.0±39.0 µg m-3, respectively. The fraction of PM2.5 over PM10 was 79%. The most important PM2.5 component was organic carbon (OC) with 37.2% of the collected PM2.5 mass. The mass fraction of OC3 in OC and EC1-OP in elemental carbon (EC) were 34.2% and 93%, respectively. Therefore, OC3 and EC1-OP can be considered as BB tracers. In addition, 68.9% of OC was water-soluble (WSOC). This implies a change of chemical properties in cloud droplets and the subsequent solar radiation extinction when the transported aerosol entering cloud layers. Levoglucosan is predominant in anhydrosugars in the near-source BB region. The ratio of chemical components of char-EC/soot-EC was found better than OC/EC in assessing BB influence and the degree of burning in this study. Moreover, the ratio of levoglucosan/mannosan was 13.3, which suggested the burning material was soft wood at the DAK site.
During the same observation period, PM2.5 and PM10 mass levels were averaged at 21.1±9.6 and 25.1±10.9 µg m-3, respectively, at Mt. Lulin. The mass fraction of PM2.5 in PM10 was 84.2%. Meanwhile, nss-SO42- and OC were the two most important PM2.5 components with the mass fractions of 14.2% and 24.0%, respectively. Moreover, the mass fraction of OC3 in OC and EC1-OP in EC were 36.5% and 77.0%, respectively. During the non-biomass burning (NBB) period, nss-SO42- was the most important component with the mass fraction of 24.8% in PM2.5. The mass fraction of OC in PM2.5 was 22.8%. OC3 was still the highest component in OC. However, EC2 became the highest component in EC. The changes in major components of PM2.5 and EC indicate the modification of contributing sources in PM2.5 during long-range transport (LRT). In addition, the ratios of PM2.5 components helped to infer that vehicle emissions and coal burning were the two major source types at Mt. Lulin during the NBB period. Since nss-K+, NO3-, levoglucosan, and di-acids and their salts were enhanced at the downwind Mt. Lulin site compared between the BB and NBB periods, they can be considered as tracers for transported BB plume.
By inspecting all PM2.5 component fractions, nss-K+ was found a stable component with the mass fraction of 2% in PM2.5 in the upwind and downwind sites. A Modification Factor (MF) was adopted to determine enhancement or degradation during LRT. The results showed that nss-SO42- was enhanced mostly indicating nss-SO42- was either enhanced by joining the existing nss-SO42- aerosol or converting from its precursor gas in the BB plume during LRT. The degradation of levoglucosan was the greatest which implies its disappearance in the atmosphere for staying a long time.
Given the fact that the quartz fiber filters used in collecting PM2.5 carbonaceous components will adsorb volatile organic compounds. OC1 of carbonaceous components was found interfered mostly in this study. The deviations of interference were significant when PM2.5 mass concentration was lower than 77 µg m-3. Even in the transported BB plume, OC1 showed significant deviation when compared two or three filters in series used in aerosol collection with one filter. In water soluble-inorganic ions, Cl- was the one with greatest volatility during collection. The corrected over non-corrected ratios of Cl- in the near-BB sources and transported BB plume were 3 and 2.2 folds, respectively.
Four source types were resolved from near-source BB aerosol using Positive Matrix Factorization (PMF). Seventy two percent of PM2.5 mass concentrations were associated with BB at Mt. DAK. The results are consistent with the classification of backward trajectory analysis. Nonetheless, anthropogenic and sea salt still show their influences even in the near-source BB area. Six source types were resolved from PMF with 77.8% of PM2.5 mass concentrations with BB origin at Mt. Lulin. It confirms the fact of LRT of BB plume. |
