欲了解中高層大氣之結構與運動模式，我們可以自離子的光化學反應著手進行探討。分布於中高層大氣F層的氧原子離子於能階躍遷的過程中，釋放出的可見光波段為630.0奈米，我們利用團隊共同架設之光度計(Photometer)系統，於夜晚的鹿林山天文台(23.46°N, 120.87°E)觀測台灣上空之氣輝，每10分鐘自光子計數器取一筆平均數據，以及俄羅斯伊爾庫茨克(51.8°N, 103.1°E)日地物理研究所之2016全年夜間氣輝地面觀測資料。並參考Link和Cogger. (1988)、Sobral et al. (1993)，以及Vladislav et al. (2008) 的現有理論，建立光化學模型進行反演，可得氧原子濃度隨高度與時間的改變，並且與福衛三號之衛星觀測資料進行比對，以此方法進行長期觀測後，了解全年變化模式。於本篇論文中，吾人採用由Solomon (2017) 所開發之氣輝模型GLOW第0.98版，產生以瑞利為單位的氣輝亮度，結合GLOW本身使用的IRI-90背景參數，來驗證此三個模型的逆向推導效果。
本團隊所演算之氧原子離子密度變化及趨勢，與福衛三號電子密度觀測結果、地面站觀測之氣輝輻射率，以及GLOW模型輸入之變數進行比較，其中異同均於本文中被討論。本團隊研發之逆推模型所解析的氧原子離子變化趨勢，在不遠的未來可望被使用於更廣泛的電離層組成變化分析。 ;To study the chemistry and composition of the upper atmosphere, we can utilize airglow emissions from the photochemical reactions of the ions in this region. When the atomic oxygen ions distributed in the ionospheric F region experience an energy level transition, visible light with a wavelength of 630.0 nm is released. We used the photometer system built by our team at NCU to perform ground-based observations of airglow over the sky of Taiwan at Lulin Observatory (23.46°N, 120.87°E) during selected night times. Ground-based airglow spectrometer observations throughout 2016 from the Institute of Solar-Terrestrial Physics (ISTP) in Irkutsk, Russia (51.8°N, 103.1°E) are also utilized. [22] We combined the mean values of our observations every 10 minutes with photochemical models based on the formulas derived from the theories of Link and Cogger (1988), Sobral et al. (1993), and Vladislav et al. (2008). With these different methods, we can estimate how the density of oxygen atomic ions varies with time and altitude and compare the results from empirical models with satellite-based observation data from FORMOSAT-3/COSMIC. The airglow brightness values simulated (Unit: volume emission rate) by the empirical GLOW model v0.98 by Solomon (2017) are also applied to validate the effectiveness of the three inversion models used in this research. The tendency and variation of the atomic oxygen ion density calculated by our photochemical models is compared to the ground-based time variation of airglow radiance, electron density observations of FORMOSAT-3/COSMIC, and input variables from GLOW. Similarities and differences are discussed. The pattern of atomic oxygen ion variation resolved by our inversion model will be utilized for further analysis of ionospheric composition variation in the future.
