| 摘要: | 大氣氣膠是估算地球輻射驅動力最大的不確定性來源，具有極端吸收能力的煤煙與其他物質混合後，顯著的輻射特性改變是可能的重要因素，在全球暖化效應的評估有顯著的影響。本文提出了三個面相的研究，可以了解煤煙在衛星遙測反演氣膠參數的影響與重要性。首先，是接續Lin等人 (2013) 探討沙塵與煤煙混合效應的研究。Lin考慮到不同沙塵特性、煤煙混合權重已計算得到包括衛星顯反射率、氣膠光學厚度 (aerosol optical depth, AOD)、混合權重與單次散射反照率 (single scattering albedo, SSA) 等變量的資料庫。本文藉由AERONET (aerosol robotic network) 長期的觀測資料決定沙塵特性以建立查找表應用於衛星資料的反演。透過個案分析顯示，在考慮沙塵-煤煙混合後輻射參數之變化後，AOD的反演誤差可從30%的低估改進為6%左右，且合理地提供煤煙混合權重與氣膠參數的空間分布。另外一方面，目前大部分的衛星反演法都有一個很基本的限制在於需要事先得知或是假設氣膠的種類以便後續的反演。這之中，煤煙成分很少被提出來考慮。因此，我們考慮到煤煙成分的重要性提出了一個新的衛星反演法—同時輻射率定法，以反演AOD與SSA。其概念為假設某適當大小視窗內的大氣條件相同，回歸分析輻射傳送方程以求解大氣的總穿透率與反射率。同時，以強吸收 (含煤煙) 與強散射的氣膠種類解析AOD在總穿透率的可能範圍，配合真實情況的大氣反射率反演AOD與SSA。目前，此法應用於地球同步衛星向日葵8號衛星上確實能偵測出台灣主要的汙染熱區。與AERONET觀測比較，其AOD反演誤差多在20%以內，顯示同時輻射率定法在氣膠參數反演之準確性及可行性。最後，在大多數情況下，大氣中的氣膠容易與水氣交互作用造成吸濕性的粒徑生長會造成散射能力的改變。因此，對於衛星反演，相對濕度 (relative humidity, RH) 是另外一個必須考慮的因子。為探究相對濕度對於煤煙光學特性的影響，本文在第三個部分根據Fan等人 (2014) 所歸納的煤煙吸濕生長因子，使用團簇‐團簇聚集模型 (cluster-cluster aggregation, CCA) 與廣義多顆粒米氏解 (generalized multi-sphere Mie-solution, GMM) 建構煤煙聚集物的樣本並分別計算煤煙在各個相對濕度下的輻射參數，再利用大氣層頂反射率的模擬評估煤煙的吸濕效應。透過相對濕度資料與中解析度影像頻譜輻射儀 (moderate resolution imaging spectroradiometer, MODIS) AOD產品的比較：霧霾事件在RH<65%、RH=65%~75%和RH>75%三種情況下平均的反演誤差分別為+42.42%、-1.80%和-33.15%。整體研究成果在氣膠輻射參數之反演具有顯著之成效，尤其是克服氣膠混合效應之影響，對於衛星監測大氣氣膠之應用有相當的貢獻。
;Atmospheric aerosols are the largest source of uncertainty in estimating the radiative forcing of the earth. When mixed with other substances, soot, a substance with extreme absorption capacity, can significantly change the radiation characteristics of aerosols, therefore imposes a significant impact on assessment of global warming effects. This study presented three aspects of research, which enabled understanding of the influence and importance of soot in retrieving aerosol parameters of satellite remote sensing. First, following the research conducted by Lin et al. (2013), where the mixing effect of dust and soot were investigated, and databases of variables such as satellite apparent reflectance, aerosol optical depth (AOD), mixing weights and single scattering albedo (SSA) were obtained through the consideration of different dust characteristics and soot mixing weights. This study determined dust characteristics using the long-term observation data of AERONET (aerosol robotic network) to create a table that can be used to retrieve satellite data. Case analysis showed that by taking into account the change of radiation parameters after dust was mixed with soot, the AOD retrieval error could be improved from underestimated 30% to about 6%, also, a reasonable space distribution of soot mixing weights and aerosol parameters could be obtained. On the other hand, currently, most satellite retrieval methods have a very basic limitation that the type of aerosol is required to be known or assumed in advance to facilitate the subsequent retrieval, with soot composition rarely proposed for consideration. Therefore, considering the importance of soot composition, we proposed a new satellite retrieval method—simultaneous radiation solution—to retrieve AOD and SSA. The concept was built on the assumption that the atmospheric conditions in a properly sized window were the same, with regression analysis of the radiative transfer equation performed to solve the total transmittance and reflectance of the atmosphere. At the same time, the possible range of AOD under certain total transmittance value was analyzed using aerosols with strong absorption (soot-containing) and aerosols with strong scattering to retrieve AOD and SSA based on the real atmosphere reflectance. At present, applying this method to the Himawari-8 geostationary satellite can indeed detect the main pollution hotspots in Taiwan. Compared with the AERONET observation, the AOD retrieval error is mostly within 20%, indicating the accuracy and feasibility of using the simultaneous radiation solution in aerosol parameter retrieval. Finally, in most cases, the aerosols in the atmosphere were likely to interact with water vapor, causing the growth of hygroscopic particle size, which changed the scattering ability. Therefore, for satellite retrieval, relative humidity (RH) is another factor that should be considered. Based on soot hygroscopic growth factors summarized by Fan et al. (2014), in the third part of this study, a cluster-cluster aggregation (CCA) model and the generalized multi-sphere Mie-solution (GMM) were used to construct samples of soot aggregates, with the radiation parameters of soot at each relative humidity calculated to explore the influence of relative humidity on soot optical properties. Then the simulation of the top-of-atmosphere reflectance was employed to evaluate the hygroscopic effect of soot. Through the comparison of relative humidity data and moderate resolution imaging spectroradiometer (MODIS) AOD product, it can be seen that: the average retrieval error in haze determination for RH<65%, RH=65-75% and RH>75% was +42.42%, -1.80% and -33.15%, respectively. The overall research results achieved significant effects in the retrieval of aerosol parameters, especially in the influence of overcoming aerosol mixing effects, making considerable contributions to the application of satellite monitoring of atmospheric aerosols. |
