| 摘要: | 黃沙現象是東亞地區春季相當活躍的現象，每年春季在中國西北沙漠地區產生的沙塵暴，經由天氣系統的傳送而影響到台灣。台灣地區在受到大陸沙塵的影響期間，大氣中的微粒濃度值明顯增加，不僅影響台灣的空氣品質，造成能見度衰竭，並影響人類的健康。為了瞭解大陸沙塵對於都會區氣膠特性的影響，本研究選擇經常會受到大陸沙塵影響的三月?五月間，在台北都會區進行PM10及PM2.5氣膠的監測。除利用環保署超級測站多種自動監測儀器量測氣膠的物化特性外，並配合人工採樣進行氣膠採樣及分析，以期獲得北部地區在黃沙時期較完整的PM10及PM2.5氣膠的物化特性。 今年襲台的沙塵暴共有八波，從PM10即時監測質量濃度值來看，影響最為嚴重的黃沙事件為第二波(3月6日?9日)及第三波(3月18日?20日)，PM10最大小時濃度值可高達160μg/m3以上，而這八波黃沙事件時期，PM2.5及PM10氣膠的平均濃度為28.7及66.3μg/m3，比起平常日PM2.5的37.4及PM10的濃度值55.4μg/m3濃度值，發現PM10的濃度值有增加，並且是粗粒徑氣膠(PM10-2.5)部分增加最多。將PM10-2.5的逐時濃度變化配合風向風速變化看來，發現當PM10-2.5濃度大於PM2.5濃度時，此時的風向明顯從原來的東北風轉成北風，且風速漸漸上升從平常日約為0.4m/sec上升最大可達1.4m/sec，所以利用PM10-2.5的逐時濃度與風向風速的配合可以推斷當地受到大陸沙塵影響的起迄時間。而從PM2.5氣膠的主要化學成份來看，硫酸鹽、硝酸鹽、碳成份濃度在黃沙時期的平均值低於平常日的濃度，顯示由沙塵暴帶來的PM2.5氣膠污染屬於少量，都會區細粒徑氣膠的貢獻還是以當地產生居多。而PM10氣膠體積濃度粒徑分佈在非黃沙時期以細粒徑氣膠居多，在黃沙時期轉變成以粗粒徑氣膠居多。 在黃沙時期進行人工的採樣，採樣分析結果在PM10及PM2.5質量濃度及碳成份的變化趨勢均與自動即時監測結果一致，顯示即時監測值的可性度。將人工採樣樣本依據逆溯氣流軌跡線可分成六類，屬於黃沙時期的逆溯氣流軌跡，其樣本在質量濃度的表現上也是以PM10-2.5居多，而非黃沙時期的逆溯氣流軌跡樣本中，以PM2.5居多，但隨著氣流經過不同的區域，其粗細粒徑氣膠化學性質也會有所不同。以碳成份來看，無論樣本是屬於黃沙時期與非黃沙時期的逆溯氣流軌跡，其濃度值變化不大。在黃沙時期，明顯從大陸沙塵源區直接傳輸過來的氣流，粗粒徑氣膠的元素組成以Si、Ca、Al、Fe居多，並且利用加強因子法發現這些元素由塵土所貢獻。同樣的在黃沙時期的逆溯氣流軌跡中，Ca2+、Mg2+所佔的比例明顯增高。利用海水加強因子法配合逆溯氣流軌跡顯示，Cl－、Mg2＋等離子來自海水飛沫，而Ca2+、SO42-、K＋來自非海水飛沫。 Yellow sand (YS) phenomenon is very active in springtime of East Asia. For the right atmospheric condition, the dust storm from northwestern region of China, the desert area, will transport dusts to Taiwan. During the period effected by the dust storm, the aerosol concentration in Taiwan increases significantly. Not only the air quality, but also the human health is threatened. Therefore, it is important to understand the effects of YS to our atmosphere. In this study, the concentrations of PM10 and PM2.5 in metropolitan Taipei from March to May in 2002 were monitored. In addition to automatic instruments installed in EPA aerosol supersite, manual sampling equipments are employed to get more detailed information of both chemical and physical characteristics of aerosols. This year, the YS invaded Taiwan eight times. From the real-time monitoring concentrations of PM10, the worst two batches were the second (March 6-9) and the third (March 18- 20) ones. The maximum hourly PM10 concentration was higher than 160 mg/m3. Among the eight batches of YS, the average of PM2.5 and PM10 was 28.7 and 66.3 mg/m3, respectively. In contrast to 37.4 and 55.4 mg/m3, the average of PM2.5 and PM10 during non-yellow-sand (NYS) periods, the concentration of the coarse particles (PM10-2.5) increased a great amount during the YS periods. It is found that on the arrival of a YS batch, the wind direction shifted from northeast to from north and the wind speed was lifted. Meanwhile, the PM10-2.5 level was increased during this time period, which is different from the level in the NYS periods. This demonstrates that one may use the change of WD and WS to infer the arrival of a YS event. Certainly, this inference is better verified by the change of the trend of PM10-2.5 level. From the chemical compositions of the aerosols, it could be found that the average mass fractions of sulfate, nitrate and carbon are lower in YS periods than that in NYS periods. It demonstrates that the dust storm brings little PM2.5, most PM2.5 of is contributed from local activities. In addition, the PM10 volume size distributions show that fine fraction is predominant in NYS periods, while coarse mode is more significant in NY periods. During NY periods, PM2.5 and PM10 were also collected manually, the results showed a consistency in variations of mass and carbonaceous content with automatic continuous measurements. This demonstrates the reliability of the manual collection as well as the automatic continuous measurement. From the HYSPLIT model (Draxel, 1999), one can categorize the 72-hour backward air trajectory into 6 types. For the periods associated with the YS backward air trajectory, the PM10-2.5 level is higher than PM2.5 level. The trend is reversed for the NYS periods. For aerosol carbonaceous contents, the variation between YS and NYS is very little. As to the elemental contents, the air trajectory associated with source regions carry predominant Si, Ca, Al, and Fe in the coarse particles; the enhancement factor calculation shows these elements were contributed from crustal materials. For water-soluble ions, the air trajectories from source regions transport predominant Ca2+ and Mg2+. From the application of enhancement factor to the water-soluble ions, Cl－and Mg2＋were originated from the sea, while Ca2+, SO42-, and K＋ were from non-sea-salt sources. |
