太陽輻射為驅動地球系統中自然現象的主要能量。無論是在大尺度的氣候模式中,森林區域的蒸發散估計,甚至在小尺度林冠層的碳通量研究上,向下地表太陽輻射通量(Downward Solar Irradiance,以下簡稱DSI)都扮演了主要的角色。過去許多研究嘗試發展進行複雜地形的輻射模式,試圖精確掌握DSI於地表之分佈。然而,台灣地區對於DSI的研究主要在於利用觀測資料來分析氣膠之效應,僅賴(2003)利用GMS-5衛星資料估算陳有蘭溪集水區內太陽輻射量之時空。本研究選擇Satellite-Based DSI Estimation Model(SDEM)來進行台灣地區DSI的推估,並考量到台灣地形的起伏,配合使用DTM地形資料,希望能夠利用地球同步衛星資料來推估DSI在台灣地區複雜地形上的時空分布。 本研究將太陽輻射傳遞過程中的主要影響因子區分為天文因素、大氣效應和地形效應,進行影響DSI程度的探討並且試圖以經驗式表示,設計出一套模式來進行地面DSI估算。2006至2007期間的晴空個案之SDEM估算值與實際地面觀測的分鐘平均值比對,兩者相關係數皆在0.93以上,有雲個案之相關係數則為0.81以上。SDEM估算值與中央氣象局地面測站觀測之DSI小時累積量比對,低海拔的測站之相關係數均超過0.9,高海拔的測站雖然因為衛星影像辨識霧的能力不佳,但SDEM估算值與高海拔測站比對結果之相關係數仍有0.8以上。 本研究SDEM模式目前僅採用MTSAT衛星可見光波段資料進行大氣效應的修正,未能充分反映大氣效應的作用,未來可在SDEM模式中引進更多紅外光波段的資料,以提高大氣效應修正經驗式之準確性。
Solar radiation is the primary driving force in global system. Downward Solar Irradiance (DSI) plays an important role in synoptic climate model, mesoscale evapotranspiration, and microscale canopy flux. Many studies tried to develop radiation models to compute DSI components on rugged terrain. However, little research has been done on the estimation of DSI over Taiwan. The purpose of this study is to estimate DSI and determine DSI distribution over the complex terrain of Taiwan. A model which based on satellite images and digital terrain data, called SDEM ( Satellite-based DSI Estimation Model), was developed. The SDEM includes the factors from astronomy, atmosphere, and terrain effects. The DSI estimation from SDEM comparing with minutely DSI measurement at RCEC Tainan and Taipei site showed the correlation coefficients were above 0.93 in clear sky cases and 0.80 in cloudy cases during 2006 to 2007. The validation of SDEM estimation with hourly accumulated measurement at 20 sites of Central Weather Bureau (CWB) showed the correlation coefficients were above 0.90 at the sites which are located in plain areas. Although the MTSAT visible band images provide inadequate evidences to detect fog in mountain region, the correlation coefficients of hourly accumulated value at CWB mountain sites were still above 0.8. The SDEM in this study only considers the MTSAT visible band images to compute the atmospheric effect. The results showed the parameters were unable to describe the detail of atmospheric properties. We expect IR-channel images used in SDEM to calculate the atmospheric effect could improve the estimation accuracy of DSI in the future.