複雜地形的向下太陽輻射通量(Downward Solar Irradiance，DSI) 之估算，有助於地球生物圈各種民生活動以及科學探索；透過地球同步衛星高時空解析度影像資訊，則是估算地面DSI時空間演變的最佳工具。臺灣地區過去對於DSI 的研究，尚止於利用零星的地面DSI 觀測資料來分析大氣懸浮微粒之輻射衰減效應，本研究沿用賴(2003)之方法，將陳有蘭溪集水區DSI 估算方法擴大到臺灣全島，這一DSI估算流程(Satellite-Based DSI Estimation Model，簡稱為SDEM)是結合高解析度的MTSAT 衛星可見光影像資料和臺灣地區數位地形資料，考慮大氣懸浮微粒、雲以及地形斜坡的遮蔽與漫射反射效應等，最後推算得出MTSAT衛星每一像元(pixel)上的DSI數值。研究結果發現，由晴空個案之下的SDEM 估算值與中央研究院環境變遷中心DSI 地面觀測值進行比對，兩者相關係數可達0.93 以上，有雲個案之相關係數也達0.81 以上。SDEM 估算值與中央氣象局氣象站觀測之DSI 小時累積量比對，在低海拔的測站之相關係數可超過0.9，與高海拔測站的相關係數稍低(~0.8)，MTSAT pixel解析度以及次網格的雲霧都是造成SDEM估算DSI失準的原因。然而，SDEM提供了大範圍逐時的DSI分布和變化，可以滿足環境調查、太陽能源應用與生態研究等應用之參考。

The estimates of Downward Solar Irradiance (DSI) of the complex terrain, contribute to the understandings of various livelihood activities and scientific explorations in the Earth's biosphere. By taking advantages of the high spatial and temporal resolution, a geostationary satellite image becomes the best tool to estimate the distribution of DSI.The former studies on DSI in Taiwan were limited to the use of ground DSI sporadic observation data to analyze the radiant attenuation of atmospheric aerosols. In this study, we followed Lai’s (2003) work over Chenyoulan River watershed and extended the domain to the whole Taiwan region. The DSI estimation process (Satellite-Based DSI Estimation Model, referred to as SDEM) is a combination of the MTSAT satellite high resolution visible image data and 40m resolution of digital terrain data in Taiwan. SDEM
considers the effect of atmospheric aerosols and clouds, the terrain slopes blocking-diffusion-reflection effect,to calculate the DSI values on each pixel point of the MTSAT satellite data.The correlation coefficient between DSI estimations of SDEM and minute-resolution ground DSI observations at Taipei and Tainan cities could reach 0.81 (cloudy situation) and more than 0.93 (clear-sky situation). Comparing with the hourly-accumulated DSI of Central Weather Bureau stations’ measurements
around Taiwan, the correlation coefficients are 0.8 (high-altitude station) to more than 0.9 (low-altitude stations). The deviation of SDEM DSI estimation might be caused by the coarse resolution of MTSAT pixel
resolution and the sub-grid cloud and fog effect. Nevertheless, the large spatial region with hourly resolution of DSI computed from SDEM could provide useful information for environmental investigations, solar energy
application and ecological researches in Taiwan.