| 摘要: | 氣候異常導致台灣近年來發生乾旱及淹水事件越發頻繁，造成了農村中的作物經濟損失，國際上也提出「以自然為本的解方」(Nature-based Solution, NbS)措施來應對災害，許多研究指出不同規模的NbS能夠更有效地解決地區的洪災問題，因此本研究選定桃園市大園區溪海里做為研究地區並以SWMM（Storm Water Management Model）來模擬兩種規模的設施，大型NbS是以水利署提出的加高農地田梗方式「在地滯洪」作為對應洪水之手段，小型NbS則是以綠屋頂作為施作手段。在滯洪模擬方面，以前人研究來檢定及驗證綠屋頂的參數，降雨選擇小時尺度事件，以淹水潛勢圖來檢定及驗證農地滯洪結果，降雨選擇5、10及20年重現期的設計降雨，兩者的驗證結果判定係數R2分別為0.96及0.98，表示為高度相關。選定8種情境進行效用分析，分別為田梗加高10、20及30公分各自包含有無安裝綠屋頂2種情境的組合，最後用滯洪體積比作為指標結合4個燈號來表示淹水的嚴重程度，在結果中發現加入綠屋頂只針對降雨前期有小幅滯洪作用，約為4mm之淹水，而在地滯洪在10年重現期的強度以上時可依照田梗加高10、20及30公分分別滯洪59mm、130mm及200mm的淹水，並且在實施田梗加高30公分時能夠將研究地區的淹水防護標準由5年提升至20年。在供水方面，本研究設計供水機制以判斷供水及儲水時機，假設滯洪後以總灌溉水量的3%做為儲水上限，並供水田灌溉用水，降雨選用2009~2022年日降雨，在結果中發現大量中小型降雨情況農地滯洪機制可收集之水量最多，而短延時強降雨收集到的水量較少，綜合來說，降雨時間點相對於稻米不同生長階段中水田湛水深需求對水田供水效用影響甚鉅。;Due to abnormal climate conditions, Taiwan has experienced increasingly frequent droughts and floods in recent years, resulting in significant economic losses for crops in rural areas. This study focuses on the Xi Hai village in Dayuan District, Taoyuan City, and utilizes the Storm Water Management Model (SWMM) to simulate two types of Nature-based Solutions (NbS) facilities. The large-scale NbS approach involves raising the agricultural land′s ridge elevation as a means to detain floodwater, in accordance with the Water Resources Agency′s recommendation. On the other hand, the small-scale NbS approach utilizes green roofs. The parameters of the green roof were calibrated and validated using hourly rainfall events, and the flood potential maps were used to verify the results with design rainfall events of 5, 10, and 20-year return periods. The simulations showed a strong correlation with the validation results, with R2 values of 0.96 and 0.98 for the green roof and on-site flood detention, respectively. Utility analysis was conducted for eight scenarios, which combined different ridge elevations (10 cm, 20 cm, and 30 cm) with and without green roofs. The flood volume ratio was used as an indicator, along with four color-coded indicators, to evaluate the severity of flooding. The results indicated that green roofs had a minor detention effect during the initial rainfall, reducing flooding by approximately 4 mm. On the other hand, on-site flood detention proved effective, with ridge elevations of 10 cm, 20 cm, and 30 cm detaining floodwater by 59 mm, 130 mm, and 200 mm, respectively, for rainfall events exceeding the 10-year return period. By implementing a 30 cm ridge elevation, the flood protection standard in the study area could be improved from a 5-year to a 20-year return period. Regarding water supply, a mechanism was designed to determine the timing of supply and storage. After flood detention, the storage capacity was limited to 3% of the total irrigation water volume for paddy field irrigation. Daily rainfall data from 2009 to 2022 were analyzed, showing that the agricultural land detention mechanism collected the highest amount of water during moderate to small-scale rainfall events, while less water was collected during short-duration heavy rainfall events. Overall, the timing of rainfall relative to the water depth requirements during different stages of rice growth significantly affected water supply effectiveness in paddy fields. |
