2011-08-012024-05-18https://scholars.lib.ntu.edu.tw/handle/123456789/707579摘要：近年來因台中地區之快速開發與都市化發展，大大地改變該流域中原有之土地使用型態與地貌，導致水文環境發生變化而造成特殊的水文循環特性，使得模擬流域之降雨—逕流過程的困難度大為增加，也因而降低水文模擬結果的效率與可信程度。另一方面，由於全球暖化日益明顯所導致之自然氣候變異與趨向越來越極端的降雨型態變化，也衝擊著流域的水文循環，更進一步增加洪災防治的難度與不確定性。因此，探討氣候變異與都市化作用對整體流域水文循環所造成之衝擊，實是永續防災研究中一個重要的課題。故本計畫將以涵蓋台中都會區之整個烏溪流域為研究對象，探討氣候變異與都市化前後（以1999年為分界點）對烏溪流域中降雨特性與地表水文特性之影響。 在研究項目上，將使用近30年間（由1980至2009年）的烏溪流域中歷史重大颱風事件的降雨—逕流資料之研析成果，配合以類神經網路建置的颱風降雨—逕流模式，評估因都市化作用情境下對於地表水文特性之影響與衝擊。在模式建置過程中，將使用過去多場歷史颱洪事件進行模式之參數檢定工作。待檢定工作完成後，再以97年發生之卡玫基颱風作為驗證事件，以確定本計畫建置的模式在極端事件之模擬能力與適用性。接著，為了探討氣候變異影響下台中地區降雨型態的變化情況，將分析氣候變異前後極端降雨量與降雨型態的改變趨勢。然後，進一步依據分析結果探討烏溪流域因受氣候變異與都市化之雙重影響情境下，所造成之流域逕流量之變異情況。最後，利用已建置之降雨—逕流模式推估烏溪流域受氣候變異與都市化雙重衝擊下之逕流量，並將針對氣候變異與都市化對烏溪流域中地表水文的衝擊影響，研擬可行之調適策&#63862;與建議，供未&#63789;災害防治工作相關作業應用之&#63851;考。 <br> Abstract: Due to the development and urbanization of Tai-Chung area in recent years, the basin characteristics are changed, which leads to the change of the rainfall-runoff process in the Wu-Creek Basin. Consequently, modeling the rainfall-runoff process becomes more difficult and that may reduce the reliance on modeling results. Moreover, the climate variation and the extreme rainfall pattern, which are caused by global warming, also have influence on hydrological processes and that would result in increasing uncertainty of modeling results. Hence, assessing the impact of climate variation and urbanization on hydrological processes is important in hazards mitigation and sustainable development programs. In this project, the characteristics of rainfall and runoff data before and after 1999 are analyzed to assess the impact of climate variation and urbanization on rainfall-runoff processes of the Wu-Creek Basin. Firstly, the variation of maximum rainfall and maximum runoff in recent 30 years (from 1980 to 2009) will be calculated. The frequency of extreme events will also be analyzed. Secondly, in order to assess the impact of urbanization on the characteristics of surface runoff, artificial neural networks (ANNs) based typhoon rainfall-runoff models will be used to simulate the rainfall-runoff process of the Wu-Creek Basin. During the construction of ANN-based models, historical typhoon events are used for calibration, and Typhoon Kalmaegi is used for validation. Then, to assess the influence of climate variation on rainfall, the frequency analysis will be performed on rainfall data before and after 1999. Finally, the influence of both climate variation and urbanization on rainfall-runoff processes will be evaluated. The runoff hydrographs under the influence of climate variation and urbanization will be estimated using the ANN-based rainfall-runoff models. According to the results of this project, comments and strategies will be drafted for future disaster mitigation.氣候變異都市化烏溪流域降雨特性地表水文類神經網路Climate variationUrbanizationWu-Creek BasinRainfallSurface runoffArtificial neural networks