指導教授：徐年盛臺灣大學：土木工程學研究所謝旻希Hsieh, Min-HsiMin-HsiHsieh2014-11-252018-07-092014-11-252018-07-092014http://ntur.lib.ntu.edu.tw//handle/246246/260777　　本研究旨在以國內外區域性地下水流數值模式建置的率定方法為基礎，進而加以補充及系統化，以建立一套數值模式率定方法，並將已建立的率定方法實際應用於花蓮溪流域縱谷區，藉以檢視該方法的可行性。 　　本研究所建立之方法，主要包含地下水系統入出流量辨識、穩態水流數值模式參數檢定與暫態水流數值模式參數檢定等三部分。為充分瞭解區域地下水系統之水流進出特性，本方法首先透過地下水歷線法及氧同位素示蹤法，估算研究區域內地下水平衡與各補注源水量，以達成系統入出流量辨識；其次，藉由穩態模擬以檢定補注量、出流量與水力傳導係數之空間分布；最後，透過暫態模擬以檢視儲水係數與比出水率之空間分布。上述過程經由迭代修正，直至校正後的分區S值及Sy值與前次修正值相符為止。 　　本研究將所建立之方法實際應用於花蓮溪流域縱谷區之地下水系統。系統入出水量辨識結果顯示，花蓮溪流域縱谷區在2008至2012年間，年平均補注量為7.87億噸，其中，雨水補注佔11.70%、河水補注佔63.53%、區外地下水補注佔24.77%；年平均出流量為8.13億噸，其中，抽水佔7.03%、沿海流失佔31.23%、出滲至河川佔61.74%；年平均蓄水變化量為-0.26億噸。 　　花蓮溪流域縱谷區地下水系統之水流數值模式率定結果顯示，年平均補注量8.04億噸，其中，11.92%來自雨水、67.03%來自河水、21.05%來自區外地下水補注，總補注量與系統入出水量辨識計算結果之差異為2.16%；年平均出流量為7.73億噸，其中，7.39%來自穩定抽水、30.46%來自穩定沿海流失、62.15%出滲至河川，總出流量與系統入出水量辨識計算結果之差異為4.92%；年平均蓄水變化量為0.30億噸。南華、吳全、大榮、長橋與大富等五口地下水位觀測井之均方根誤差分別可達2.52公尺、0.29公尺、1.65公尺、2.90公尺及2.57公尺，且水力傳導係數及比出水率率定前後差值皆在1個數量級以內。 　　結果顯示，本研究建立之率定方法可初步模擬花蓮溪流域縱谷區之地下水流情勢，未來可作為各種情境模擬與分析之參考依據。　In order to properly manage the groundwater resources, we need a well calibrated numerical model to assess the effectiveness of various management strategies. 　This study aimed to be based on the methodology of numerical model calibration for regional groundwater flow system at home and abroad, then be supplemented and systematized to establish a set of numerical model calibration method, and applied which on Hualien River Basin in East Rift Valley for surveying its feasibility. 　The methodology proposed in this study is composed of 3 phases, including the identification of source/sink of groundwater system, parameter testing of a steady numerical model and parameter testing of a transient numerical model. First of all, to understand the feature of flow of regional groundwater system, the methodology utilized groundwater hydrograph analysis and oxygen isotopic tracer method to estimate the water balance and the recharge sources in study area. Secondly, the spatial distribution of source, sink and hydraulic conductivity were checked in steady-state simulation separately. Finally, the spatial distribution of storage coefficient and specific yield were checked in transient simulation. The process was iteratively corrected until the subdivision of storage coefficient and specific yield were fitted the previous corrected value. 　The results of the identification of source/sink of groundwater system practiced on the case of Hualien River Basin in East Rift Valley between 2008 to 2012 showed that, its annual average amount of recharge was 787 million tons ( recharge of precipitation accounted for 11.70%, recharge of river represented 63.53%, and the outer zone of groundwater recharge occupied 24.77%), the annual average loss amount was 813 million tons (pumping took up 7.03%, coastal loss constituted of 31.23%, and the seepage to the river stood at 61.74%) and the average annual storage changes was -26 million tons. 　The results of the numerical model of regional groundwater flow system used in the case of the daily model of Hualien River Basin in East Rift Valley indicated that, its annual average amount of recharge was 804 million tons ( recharge of precipitation accounted for 11.92%, recharge of river represented 67.03%, the outer zone of groundwater recharge occupied 21.05%, and the number difference between the total amount of recharge and the results of the system identification was 2.16%), the annual average loss amount was 773 million tons (pumping took up 7.39%, coastal loss constituted of 30.46%, the seepage to the river stood at 62.15%, and the number difference between the total amount of the outflow and the results of the system identification was 4.92%) and the average annual storage changes was 30 million tons. The root-mean-square error of Nanhua, Wuquan, Darong, Changqiao and Dafu reached to 2.52 meters, 0.29 meters, 1.65 meters, 2.90 meters and 2.57 meters, respectively. The hydraulic conductivity and specific yield were regulated within 1 order of magnitude. 　The results showed that the calibration method established in this study could initially simulate the groundwater flow situation of Hualien River Basin in East Rift Valley. Likewise, it could be used as a reference for simulations and various analyse in the future.碩士學位論文口試委員會審定書 i 誌謝 ii 摘要 iii Abstract v 目錄 vii 圖目錄 x 表目錄 xiii 第一章 緒論 1 1.1 研究緣起 1 1.2 研究目的 2 1.3 研究流程 2 1.4 論文內容 3 第二章 文獻回顧 5 2.1 國內 5 2.2 國外 14 2.3 文獻回顧總整 15 第三章 率定方法建立 17 3.1 邊界判釋 18 3.2 地下水系統入出水量辨識 18 3.2.1 地下水平衡分析──地下水歷線法 19 3.2.2 地下水補注源分析──氧同位素示蹤法 27 3.3 地下水流數值模式參數檢定 30 3.3.1 穩態模擬 31 3.3.2 暫態模擬 42 第四章 率定方法應用──以花蓮溪流域縱谷區為例 45 4.1 研究區域概述 45 4.1.1 區域地形 48 4.1.2 區域地質 50 4.1.3 水文地質 54 4.1.4 水文特性 62 4.1.5 穩定氫氧同位素資料 74 4.1.6 土地與水資源利用現況 76 4.2 歷年花蓮地區相關研究 78 4.3 邊界位置判釋 83 4.4 花蓮溪流域縱谷區地下水系統入出水量辨識 84 4.4.1 地下水平衡分析──地下水歷線法 84 4.4.2 地下水補注源分析──氧同位素示蹤法 94 4.4.3 地下水平衡分析及補注源分析結果 95 4.5 花蓮溪流域縱谷區地下水流數值模式參數檢定 96 4.5.1 穩態模擬 96 4.5.2 暫態模擬 120 第五章 結論與建議 135 5.1 結論 135 5.2 建議 136 參考文獻 139 附錄一 MODFLOW理論及套件概述 149 附錄二 花蓮溪流域縱谷區地質鑽井柱狀圖 159 附錄三 花蓮溪流域縱谷區地質剖面圖與柵狀圖 173 附錄四 鳳林地區井群抽水對附近地下水位影響評估 17513497240 bytesapplication/pdf論文公開時間：2019/08/17論文使用權限：同意有償授權(權利金給回饋本人)花東縱谷地下水歷線氧同位素MODFLOW穩態暫態thesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/260777/1/ntu-103-R01521316-1.pdf