- 學位論文
台灣地區近地面紫外線指數之特徵探討與數值模擬
The Characteristics Analysis and Numerical Simulation of Surface Layer UV Index in Taiwan
摘要
紫外線為太陽輻射波段中，波長介於100∼400 nm之間的光線，其能量約占太陽輻射總量的10%。西元1985年，臭氧洞的發現，迫使人們正視環境議題，也促使民眾更加注意紫外線相關資訊。台灣地區紫外線預報始於1997年，並於2000年完成第四代預報系統。此預報系統主要是使用UVspec模式對晴空狀態之UVI進行模擬，並使用中央氣象局之天氣預報資料，對模擬數值進行修正。本研究將針對台灣地區近地面紫外線的各項特徵進行分析探討，並以模式對台灣地區近地面紫外線進行模擬。 台灣地區紫外線具有夏高冬低、南高於北、東高於西、山地高於平地之變化特性。紫外線變化特徵除受太陽天頂角影響，天氣系統也是左右紫外線強弱的另一項因素。為了解天氣系統之雲雨帶對紫外線的影響，我們將UVspec模式模擬出之實際觀測（Fuv）與晴空理論（Fuv （晴朗無雲））之比值Tr＝Fuv / Fuv （晴朗無雲）定訂為「穿透率」。由結果中我們發現高輻射量則具高穿透率。雲量部分：低雲量具高穿透率，高雲量具低穿透率。雲種部分：低、中雲對穿透率影響大，高雲對穿透率影響則不顯著。 此外，為有效增進目前紫外線預報準確率，我們選用MM5模式模擬出之雲量資料，輸入UVspec模式中進行模擬與修正。希望藉由輸入客觀的模擬資料，取代原先利用天氣預報資料修正晴空模擬數值之方式，減少輸出結果之誤差，增加模式品質。經測試，代入MM5雲量資料所模擬出之紫外線，經二次公式修正後，的確能減少模擬與觀測間的誤差。
並列摘要
Ultraviolet (UV) light is electromagnetic radiation with a wavelength in the range 100 nm to 400 nm. In 1985, the discovery of ozone hole in South Pole forced people to face up to environment protection and health-related effects of UV radiation. The UV forecast system in Taiwan started in 1997. This system used UVspec model to simulate Ultraviolet Index(UVI) in clear sky state, and use Central Weather Bureau (CWB) forecasting data to modify simulation data. This research would analyze characteristics of near ground UV in Taiwan, and tried to use model simulation to improve UV forecasting. There are some features of UV in Taiwan: summer is higher than winter; south region is higher than north region; east region is higher than west region; mountain area is higher than level ground. Besides zenith, weather condition is also an important factor to determine UVI value. In order to know how clouds could affect UVI value, we define transmissivity (Tr), and found out high radiation has higher Tr; low cloud amounts had higher Tr, high cloud amounts had lower Tr; and high clouds had only little effect on Tr. Moreover, for improving the UV forecasting accuracy, we used cloud data simulated from MM5 model, and inputted into Uvspec model to make UV simulation. The results showed using model cloud data with modification formula to replace old method could reduce the error and enhance forecast quality.