| 摘要: | 本研究旨在利用長期地面觀測資料(太陽光度計/AERONET)針對沙塵粒子、生質燃燒煙塵及人為污染物等不同氣膠分析其輻射參數隨波長變化之特性及其間關係之探討。結果顯示，粒徑較大的沙塵粒子其粒徑分布參數(?ngst?m exponent; AE)的觀測值(0.16)遠小於煙塵與人為污染，可用於沙塵粒子之辨識，而煙塵及人為污染則較為接近。在單次散射反照率方面，煙塵隨波長增長而減小(440nm、675nm、870nm及1020nm分別為0.87、0.86、0.83及0.83)表示其吸收能力隨波長增加而增強。由於都市地區混合較多種類氣膠特性人為污染吸收特性隨波長並無明顯差異，但在675nm有最大值，表示在675nm吸收能力最弱。而沙塵粒子輻射參數隨波長變化特性則與生質燃燒煙塵相反，吸收能力隨波長增長而減弱。因此，透過氣膠輻射參數(光學氣膠厚度、單次散射反照率等)隨波長之變化特性，可針對沙塵粒子、混合型氣膠、人為污染物及生質燃燒煙塵進行辨識。由於大範圍單次散射反照率參數之提供並不普遍，例如衛星的反演資料，因此在實際區域性的應用上有其限制性，尤其在煙塵和人為污染的辨別。故本研究將藉由大氣中氣膠濃度值(微粒物質; Particulate Matter; PM值)與氣膠光學厚度關係的建立，探討在濃度相同下生質燃燒煙塵及人為污染物之氣膠光學厚度隨波長變化特性，作為辨識之參考。研究結果顯示，在相同濃度下生質燃燒及人為污染物氣膠氣膠光學厚度隨波長變化特性不同。因此，本研究進一步分析兩者在各波段間輻射特性的差異來建立氣膠種類辨識方法。由長期資料測試結果顯示，在生質燃燒源區測站，經辨識方法判斷為生質燃燒煙塵氣膠在AE值(Mukdahn site:1.56±0.13，Pimai site:1.52±0.11)與AERONET統計生質燃燒煙塵氣膠AE值(Mukdahn site:1.61±0.07，Pimai site:1.59±0.05)的特性相當一致，且在SSA及AOD值隨波長變化特性也相同。在人為污染測站，經辨識方法判斷為人為污染物氣膠在AE值(Taipei site:1.32±0.19，Taihu site:1.24±0.23) 與AERONET統計人為污染物氣膠AE值(Taipei site:1.27±0.23，Taihu site: 1.19±0.24)非常接近，且在SSA及AOD值隨波長變化特性上也相當吻合。而在在生質燃燒源區測站，經辨識方法判斷為人為污染物氣膠在AE值(Mukdahn site:1.49±0.18，Pimai site:1.31±0.33)介於AERONET統計生質燃燒煙塵氣膠及人為污染物氣膠AE值(Mukdahn site:1.61±0.07，Taipei site:1.27±0.23)之間，應為混合型氣膠。在人為污染測站，經辨識方法判斷為生質燃燒煙塵氣膠在AE值(Taipei site:1.37±0.16，Taihu site:1.25±0.21)介於AERONET統計生質燃燒煙塵氣膠及人為污染物氣膠AE值(Mukdahn site:1.61±0.07，Taipei site:1.27±0.23)之間，但較靠近人為污染物特性，可視為混合型氣膠。因此，此辨識方法在對生質燃燒煙塵及人為污染物氣膠源區測站判斷結果與AERONET統計特性一致，未來在衛星觀測的應用方面可提供相當的助益。The objective of this study focuses on the spectral characteristics from ground based observations (AERONET) for the type discrimination of the aerosols around the source regions of Asia and Africa. The suspended particles of mineral dust, smoke plumes and anthropogenic pollutants are the interesting types.The long term measurements of the AERONET stations exhibit that the mineral dusts have the large particle size, the mean value of ?ngstr?m exponent (AE) of dust particles is about 0.16, while the smoke plumes and anthropogenic pollutants are equipped with strong absorption demonstrated by the values of Single Scattering Albedo (SSA). The SSA of smoke plumes from biomass burning decreases with the wavelength increases (440nm, 675nm, 870nm and 1020nm 0.87,0.86,0.83 and 0.83, respectively), indicating that the absorption capacity increases with the wavelength increases. For anthropogenic pollutions, the spectral SSAs are stable than the others except the weakest absorption at 675nm. Moreover, dust aerosols have the opposite character from biomass burning aerosols. That is the spectral absorption of dust aerosol decreases as the wavelength increases. The results therefore suggest that it is practical to discriminate the aerosol types by means of the spectral characteristics of AE and SSA. However, the SSA property of aerosols is not frequently available for a wide area, for instance, SSA retrievals from remotely sensed data. In order to overcome such kind of limitation, this study try to distinguish smoke and anthropogenic aerosols based on the characteristics of spectral variation in optical properties under equivalent concentration (Particulate Matters; PM). By associating with the relationship between atmospheric aerosol concentration value (Particulate Matters; PM) and aerosol optical depth, the spectral variance in aerosol optical depth of biomass burning aerosols are distinct from anthropogenic pollutants. Eventually, the validation of proposed approach in discrimination smoke from anthropogenic pollutant is verified, and the results show the agreement with the observation from AERONET implying the highly practical. For more complex aerosols, further investigations are required. |
