| 摘要: | 氣膠可透過改變地球輻射收支系統影響全球氣候，NOAA (National Oceanic and Atmospheric Administration)為量化氣膠光學特性與氣膠輻射效應，建立氣膠聯合觀測網(NOAA Federated Aerosol Network , NFAN)監測全球氣膠光學長期變化，其中鹿林山大氣背景站(Lulin Atmospheric Background Station, LABS; {23.469}^\circle N, {120.874}^\circle E, \ 海拔2862公尺)為觀測網內亞洲主要觀測站之一，位於臺灣玉山國家公園內。NOAA氣膠觀測系統雖提供先進的氣膠光學參數，但不完整的氣膠微物理與化學成分觀測，對評估氣膠輻射效應仍是挑戰。因此，本研究將基於米氏散射理論的數值模型結合LABS氣膠觀測，建立一套反演法，嘗試計算出氣膠粒徑分布與複數折射率。
透過敏感度實驗，我們發現此反演法適用於雙峰粒徑分布條件，且歸納出反演相對有效範圍：氣膠細粒徑<300nm或粗粒徑<500nm，氣膠等效折射率實部(RRI)介於1.3至2.5間與等效折射率虛部(IRI)介於10-4至1之間。進一步針對真實環境檢驗此反演法之適用性，本研究選擇2020年4月及2021年3月兩個整月進行模擬，經觀測資料所建立的氣膠種類分類法可得知，此期間LABS受沙塵與生質燃燒氣膠傳輸的影響。執行本反演法的結果顯示，生質燃燒期間氣膠等效粒徑平均值約0.77\pm0.36\mu m與氣膠折射率平均值約1.82-i0.04 (1.62-i0.00至2.02-i0.08)；沙塵期間氣膠等效粒徑平均值約1.47\pm0.63\mu m與氣膠折射率平均值約1.48-i0.01 (1.40-i0.00至1.68-i0.03)，結果說明反演法可明確描述沙塵與生質燃燒氣膠變化，亦可呈現觀測期間氣膠粒徑增長之特性。反演結果與觀測比較，雖然此反演法的結果對於趨勢變化掌握良好(r>0.9)，整體來看反演法對散射係數低估約12%、吸光係數低估約4%、質量濃度高估可達35%，且吸收係數與質量濃度的反演結果有較大不確定性。未來本反演法將可應用於不同儀器間的觀測閉合度檢驗，並有助於改善輻射傳遞模式所用的輸入資料完整性，精進氣膠輻射效應估算。
;Aerosols can alter the earth′s radiation budget and influence global climate change. To improve the understanding of aerosol optical properties (AOPs) and aerosol radiation effect (ARE), NOAA Federated Aerosol Network (NFAN) was established to monitor the mean values, spatiotemporal variability, and long-term trends of AOPs. Lulin Atmospheric Background Station (LABS, {23.469}^\circle N, {120.874}^\circle E, \ 2862\ a.m.s.l.), which is located on the top of Mt. Lulin in central Taiwan, is one of the NFAN major sites in Asia. Although the NOAA aerosol system provides the state of art AOPs measurements, the lack of aerosol microphysics and chemical components information still remains a challenge for the ARE calculations. Therefore, the current study aims to develop an inversion of aerosol optical and microphysical properties (i.e., aerosol size distribution and the aerosol refractive index (RI)) based on Mie theory and aerosol in-situ measurements at the LABS.
According to the sensitivity experiment, the inversion method appropriate for the case of bimodal aerosol size distribution. The generalizing effective coverage was the fine mode of aerosol diameter <300nm or the coarse mode diameter <500nm, the real part of the equivalent refractive index (RRI) around 1.3 to 2.5, and the image part of the equivalent refractive index (IRI) around 10-4 to 1. Further, we selected two full months of April 2020 and March 2021 for implementing the inversion. As the observational results suggested by an aerosol-type classification method, the LABS was affected by the long-range transported dust and biomass burning aerosols in these two months. Base on the LABS’s realistic inversion, the monthly mean of biomass burning AOPs was around 0.77\pm0.36\ \mu m for the aerosol equivalent diameter and around 1.82-i0.04 (1.62-i0.00 to 2.02-i0.08) for the aerosol refractive index in March 2021; The mean of dust AOPs was about 1.47\pm0.63\mu m for the equivalent diameter and around 1.48-i0.01 (1.40-i0.00 to 1.68-i0.03) for the aerosol refractive index in April 2020. The above results illustrate that the inversion method can well represent the variation of dust and biomass burning aerosol properties, and the growth of the aerosol diameter. Compared to the measurements of the results, it shows a good correlation (r>0.9), but underestimates the scattering coefficient of 12%, underestimates the absorption coefficient of 4%, and overestimates the mass concentration of 35%. A larger uncertainty (i.e., root-mean-square error) was shown for the absorption coefficient and mass concentration. In the future, our inversion could apply to the observation closure between different instruments, and help to intensify the integrity of the input data in the radiation transfer model, which can improve the ARE simulation. |
