土地與氣候變遷情境對流量之影響 -以大屯溪流域為例

Impacts of Land Use and Climate Changes Scenarios on Streamflow-A case study of Datuan Watershed

林子平(Tzu-Ping Lin)

指導教授 ： 林裕彬

國立臺灣大學/生物資源暨農學院/生物環境系統工程學研究所/碩士(2015年)

https://doi.org/10.6342/NTU.2015.11032

摘要

氣候變遷對全球水文循環有不同程度的衝擊影響，而臺灣於2014年陸續進行第五次耦合模式比較計畫之研究，提出各大氣環流模式之氣候不確定性與氣候所造成之衝擊。由於地勢陡峭促使地表逕流快速地流入海洋、乾濕季節雨量極端的現象以及氣候變遷的情況下，臺灣對於水資源的管理難關重重，因此，需要及早進行分析並做好應變措施。
本研究的目的為，評估臺灣較小集水區域之未來土地利用改變與氣候變遷情形下，對於流量造成的衝擊。於氣候變遷方面，採用IPCC(Intergovernmental Panel on Climate Change)第五次評估報告(AR5)下5個大氣環流模式(General Circulation Model, GCM)輸出的資料，並做時間降尺度至日時間單位，以作為後續SWAT水文模式輸入端資料使用；而在土地利用變遷方面，採用CLUE‐s(the Conservation of Land Use and its Effects at Small regional extent)模式模擬出近未來2020年至2039年之土地利用分布情形；最後，結合氣候變遷與土地利用模式之輸出結果，輸入至SWAT 水文模式(Soil and Water Assessment Tool model)中，模擬出近未來年之流量情形，並進行分析與建議。此外，模式參數會隨著季節、氣候及空間的變化，對模式預測之結果有極大的影響，因此，本研究在率定之前，先採用不確定性估計(Generalized Likelihood Uncertainty Estimation, GLUE)方法進行SWAT水文模式參數的選取。
在近未來內，土地利用模式結果顯示林地(+1.2%)及都市(+36.8%)都將增加，農地(-15.6%)及其他類別土地(15%)將減少；大氣環流模式結果顯示總降雨量有減少的情形，並在豐水季節增加降雨(+1.44%)，枯水季節減少降雨(-3.41%)。除此之外，於土地利用改變、氣候變遷、氣候變遷下土地利用改變等三種情境下之結果顯示，土地利用僅略微影響著流量，雨量才是影響流量的最大因素，導致氣候變遷情境與氣候變遷下土地利用改變情境之結果相似，由兩者情境結果發現，豐水期間7、8月份之流量上升，僅10月份流量減少幅度較大，但整體豐水季節流量是略為增加的(+0.12%)；而於枯水期間3、4月份之流量是減少的，整體枯水季節也呈現減少的趨勢(-5.35%)，造成豐枯水季節之流量有呈現更明顯之狀況，且加上年平均流量為減少的情形，此一結果可能造成近未來年缺水相關問題的發生。此外，本研究發現研究區域內之流量趨勢可能有往前遷移的情形發生，因高流量月份由原本9、10、11月份變為8、9、10月份，而本研究認定此一原因為雨量改變所造成的。
根據結果顯示，管理未來水資源將會是一項嚴峻的挑戰，因此，氣候變遷與土地利用變化對大屯溪水資源帶來的影響，應及早規劃以應變未來的不確定性。

關鍵字

SWAT水文模式 ； 氣候變遷模式 ； CLUE-s模式 ； 流量 ； 水資源衝擊

並列摘要

Climate change projects have various levels of impacts on hydrological cycles around the world. The impact of climate change and uncertainty of climate projections from general circulation models (GCMs) from the Coupled Model Intercomparison Project (CMIP5) which has been just be released in Taiwan, 2014. Since the streamflow run into ocean directly due to the steep terrain and the rainfall difference between wet and dry seasons is apparent; as a result, the allocation water resource reasonable is very challenge in Taiwan, particularly under climate change. The purpose of this study was to evaluate the impacts of climate and land use changes on a small watershed in Taiwan. The AR5 General Circulation Models(GCM) output data was adopted in this study and was downscaled from the monthly to the daily weather data as the input data of hydrological model such as Soil and Water Assessment Tool (SWAT) model in this study. The spatially explicit land uses change model, the Conservation of Land Use and its Effects at Small regional extent (CLUE‐s), was applied to simulate land use scenarios in 2020-2039. Combined climate and land use change scenarios were adopted as input data of the hydrological model, the SWAT model, to estimate the future streamflows. In addition, the model parameters with the variation of seasons, weather, and spaces will have an extremely influence of the model prediction. Therefore, this study applies the Generalized Likelihood Uncertainty Estimation(GLUE) methods to choose SWAT model parameters before calibration. In recent year, CLUE-s model showed the increasing forest(+1.2%) and urban(+36.8%) area, and decreasing agricultural(-15.6%) and others land(-15%); GCMs model showed the decreasing annual precipitation, increasing rainfall in wet season(+1.44%) and decreasing rainfall in dry season(-3.41%). Besides, this study had three scenario: land use change, climate change and both of land use change and climate change. SWAT model showed land use didn’t influence streamflow very much in the land use change scenario, rainfall is the main reason for streamflow. Therefore, there were similar results between climate change scenario and both of land use change and climate change scenario, and both scenario showed the increasing streamflow in July and August, much decreasing streamflow in March, April and October, little increasing streamflow on wet seasons(+0.12%), decreasing streamflow on dry season(-5.35%). Cause the difference of streamflow between wet season and dry season are also increased, and decreasing annual streamflow. It will cause the problem of water shortage in the future. In addition, the study found the high streamflow will change from month 9-11 to month 8-10 in Datuan river, and it is because of rainfall. This result indicates a more stringent challenge on the water resource management in future. Therefore, impacts on water resource caused by climate change and land use change should be considered in water resource planning for the Datuan river watershed.

並列關鍵字

SWAT ； GCM ； CLUE-s ； streamflow ； water impact

參考文獻

96. 顧嘉安. (2010). 以馬可夫鍊細胞自動機模型模擬極端洪水對都市土地利用型態之影響─ 以臺北市為例. 成功大學都市計劃學系學位論文, 1-112.

95. 羅時麒. (2005). 以系統性機率模式鑑定量化與整合生命週期評估之不確定性.

93. 蕭戎雯. (2013). 不同單元尺度對土地利用及生態系統服務模擬之影響-以大屯溪流域為例. 臺灣大學生物環境系統工程學研究所學位論文, 1-126.

62. 王彥覃. (2008). 應用自上而下與自下而上土地利用及降雨-逕流模式模擬土地利用變遷與水文量-以寶橋集水區為例. 臺灣大學生物環境系統工程學研究所學位論文, 1-109.

82. 陳虹螢. (2010). 整合土地利用與水文模式於集水區規劃管理之研究-以臺北都會區為例. 臺灣大學生物環境系統工程學研究所學位論文, 1-111.