請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/31255
標題: | 氣候變遷對台灣地區降雨沖蝕指數之影響 The Impact of Climate Change on Rainfall Erosivity in Taiwan |
作者: | Wen-Jen Yang 楊文仁 |
指導教授: | 范正成 |
關鍵字: | 氣候變遷,降雨沖蝕指數,最大三十分鐘降雨強度,氣候生成模式,類神經網路,全球暖化, Climate change,Rainfall erosivity,The maximum 30-minute rainfall intensity,Climate generator,Artifical neurual network,Global warming, |
出版年 : | 2006 |
學位: | 博士 |
摘要: | 摘 要
本研究擬蒐集中央氣象局及水利署於全台各地之雨量站資料,分析歷史降雨事件之相關統計特性,以進行CLIGEN氣候生成模式於台灣地區適用性之驗證,並以全球環流模式GISS以及HADCM3模式之輸出值,模擬分析未來氣候變遷對台灣地區降雨沖蝕指數(R指數)之影響,完成未來氣候變遷台灣地區年等降雨沖蝕指數圖之繪製。 研究結果顯示,台灣地區之有效雨量佔年平均雨量之80%以上,其最大三十分鐘降雨強度(I30)之分佈型態,西南部屬低闊峰分佈,東北部則屬於高狹峰的分佈型態,且I30與平均降雨強度有較佳的相關性。依據有效降雨事件之分析結果,建立I30迴歸公式以及倒傳遞類神經網路(BPN),而BPN在I30的推估上,可以獲致較佳的結果。此外,CLIGEN氣候生成模式對於台灣地區年雨量與月雨量之模擬結果良好,惟對有效降雨事件之重要特性的模擬,仍有待進一步之評估。 GISS平衡試驗之模擬結果顯示,於兩倍CO2濃度情境時,基隆、宜蘭、新竹、嘉義以及花蓮地區,其年平均R指數有增加之趨勢,高雄與台東地區則減少。HADCM3漸變試驗之模擬結果亦顯示,未來氣候變遷對台灣地區R指數有顯著之影響,除了B2情境短期,部分地區是減少的外,其餘無論短期、中期或長期均有增加之趨勢,尤以長期之影響為最,其中又以西南部地區之R指數受其影響為最,而對花蓮台東地區之影響則最小。若不考慮RUSLE公式中其他因子之變化,則R 指數之增加率即為土壤沖蝕之增加率;惟此一假設並未考慮氣候變遷對C因子的影響,氣候變遷對土壤沖蝕之影響仍有待進一步之評估。 ABSTRACT In this study, the rainfall data were collected from Central Weather Bureau and Water Resource Agency in Taiwan. Using these data, the CLIGEN model was validated for Taiwan based upon the attributes of historical rainfall events from the analysis results. The impact of climate change on rainfall erosivity (the R factor) in Taiwan was assessed as well. Results show that the effective rainfall in Taiwan was more than 80% of the mean annual rainfall. The distribution types of the maximum 30-minute rainfall intensity (I30) in southwest and northeast area were leptokurtosis and platykurtosis respectivily, and there was a good relationship between I30 and mean rainfall intensity. According to the analysis results of effective rainfall events, a regession model and a Back-Propagation artifical neurual network (BPN) were established, and the BPN has a better performance in predicting the I30. Apart from these, the CLIGEN weather generator was found to be satisfactory in modeling the annual precipitation and monthly precipitation in Taiwan. However, further validation for the other important attributes of effective rainfall events was required. The simulation results of the GISS equilibrium experiment show that the annual rainfall erosivity tends to increase in Keelung, Ilan, Hsinchu, Chiayi and Hualien, and decrease in Kaoshiun and Taidong under the scenario of 2 times concentration of CO2. From the results of HADCM3 transition experiment, it was found that future climate change might have significant influence on rainfall erosivity in Taiwan. Except for B2 scenario, there was an increasing trendency for the R factor during the short, medium and long term, southwest area in long term especially. The increase in the R factor will lead to increase the soil erosion as well, if other factors in the Revised Universal Soil Loss Equation (RUSLE) remain constant. However, the assumption doesn’t consider that the influence of climate change on the factor C, the research for the impact of climate change on soil erosion needs to be further studied. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/31255 |
全文授權: | 有償授權 |
顯示於系所單位: | 生物環境系統工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-95-1.pdf 目前未授權公開取用 | 2.1 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。