集水廊道附近三維地下水流數值模式率定方法建立與應用－以二峰圳為例

Establishment and Application of 3D Numerical Model Calibration Method for Groundwater Flow Near Infiltration Gallery– A Case Study of the Er-Feng Canal

鈕希婷(Xi-Ting Niu)

指導教授 ： 徐年盛

國立臺灣大學/工學院/土木工程學研究所/碩士(2018年)

https://doi.org/10.6342/NTU201703880

摘要

本研究旨在建立集水廊道附近三維地下水流數值模式率定方法並將其應用於林邊溪二峰圳以彰顯所建立方法之可用性。本方法主要分為四步驟，包含水文地質概念模式建立、地下水系統水平衡分析、地下水流數值模式建立以及數值模式率定優選模式建立。
本研究首先蒐集研究區域之地下水鑽探井、河道地質探測以及地下水位觀測井之相關資料，建立其水文地質概念模型並由水平衡分析結果了解地下水系統之流入或流出狀況，而後建立一三維飽和與未飽和地下水流數值模式以模擬集水廊道附近之地下水流況，同時參考前人研究後提出一合理且誤差在容許範圍內之粗砂未飽和土壤參數曲線之修正曲線，並使數值模式得以收斂。接著建立一套自動化數值模式率定方法進行穩態數值模式率定以獲得透水係數及未飽和土壤參數，然後以率定完成之結果為基礎，再進行暫態模式率定以獲得河道流量以及河道補注量與蒸發量之長期趨勢。最後利用率定完成的參數再重新進行一次地下水系統水平衡分析，並與數值模式結果相互驗證。
林邊溪二峰圳集水廊道附近地下水流數值模式率定結果顯示，河道沖積層之透水係數大約介於93(m/d)至360(m/d)，河床質表層透水係數大約介於2.7(m/d)至8.8(m/d)，廊道濾層則為176(m/d)，皆在其透水係數之合理範圍內。而觀測井C於2010年1月1日至2010年9月5日旱季與雨季之均方根誤差分別為0.24公尺及0.25公尺，觀測井D於旱季與雨季之誤差分別為0.22公尺及0.23公尺，觀測井E於旱季與雨季之誤差分別為0.37公尺及0.46公尺，集水廊道集水量於旱季與雨季之誤差分別為5,746(cmd)及26,743(cmd)，誤差比例分別為16%以及15%。另外，數值模式所計算之地下水系統總流入量約為50.49百萬噸，總流出量約為45.96百萬噸，而地下水系統之水平衡分析所計算之總流入量約為46.14百萬噸，總流出量約為45.25百萬噸。

關鍵字

林邊溪 ； 二峰圳 ； 集水廊道 ； FEMWATER ； 優選模式 ； 參數檢定 ； 未飽和土壤

並列摘要

The aim of this study is to establish a 3D numerical model calibration method for groundwater flow near infiltration gallery and apply to the Er-Feng Canal of Lin-Bien River to prove the availability of the method. The numerical model calibration method established by this study includes four parts: hydrogeological conceptual model, water balance estimation of groundwater system, numerical model of groundwater, and optimization model for parameter verification.
First, this study collects the drilling data, resistivity image profile and observation data near study area, and establishes a hydrogeological conceptual model to realize groundwater inflow and outflow condition by water balance estimation, and then builds a 3D saturated and unsaturated groundwater numerical model to simulate groundwater flow near infiltration gallery. Meanwhile, referring to previous results, this study proposes a reasonable curve of unsaturated soil with small error to revise the hydraulic conductivity of coarse sand, which makes the model converges. After that, this study establishes an auto-optimization model for parameter verification to get the hydraulic conductivity and the parameters of SWRC in steady-state model, and conducts parameter verification to get river recharge and evaporation in long-term trend in transition model by the results of steady-state model. Last, using the calibrated parameters to calculate water balance of groundwater system once again, and compares with the results of numerical model.
According to the results of calibrating groundwater flow model near infiltration gallery of Er-Feng Canal, the hydraulic conductivity of riverbed alluvial is between 93(m/d)~360(m/d), the river bottom sediment is between 2.7(m/d)~8.8(m/d), the filter layer of infiltration gallery is 176(m/d), all of them are in a reasonable range. Then, root mean square error (RMSE) during dry and rainy season in 2010/01/01~2010/09/05 of the observing well C are 0.24 m and 0.25 m. The observing well D are 0.22m and 0.23 m, the observing well E are 0.37 m and 0.46 m, and water intake of infiltration gallery are 5,746 cmd and 26,743 cmd respectively, the error ratio of them are 16% and 15%. In addition, the total recharge of groundwater system is 50.49 million tons and total outflow is 45.96 million tons. According to the results of water balance, the total recharge of groundwater system is 46.14 million tons and total outflow is 45.25 million tons.

並列關鍵字

Lin-Bien River ； Er-Feng Canal ； Infiltration gallery ； FEMWATER ； Optimization ； Calibration ； Unsaturated soil

參考文獻

1. Bear, J. (1979). Hydraulics of groundwater. McGraw-Hill Book Co.

2. Becker, L., and Yeh, W. W.-G. (1972). Identification of parameters in unsteady open channel flows. Water Resources Research, 8(4), 956-965.

3. Brooks, R.H., and Corey, A.T. (1964). Hydraulic properties of porous media. Hydrology Paper, 3, Colorado State University, Fort Collins, Colorado.

4. Carsel, R. F., and Parrish, R. S. (1988). Developing joint probability distributions of soil water retention characteristics. Water Resources Research, 24(5), 755-769.

5. Chen, W., Huang, C., Chang, M., Chang, P., and Lu, H. (2013). The impact of floods on infiltration rates in a disconnected stream. Water Resources Research, 49, 7887-7899.