| 摘要: | 在複雜的沖積扇含水層系統中，建立良善的異質性水文地質模型，可有效的進行地下水管理與緩解地層下陷問題，對於水文地質學、工程地質學和環境科學領域至關重要。本研究旨在闡明從鑽探資料到建立異質性水文地質模型對水文地質學的影響，目標在解決三個挑戰：(1) 量化水文地質模型不確定性對局部尺度地下水流和地層下陷模擬的影響；(2) 評估鑽孔密度對盆地尺度三維異質性水文地質空間分布特性與地質模型建構的影響；(3) 發展異質性水文地質模型下之地下水和地層下陷的非耦合模型，以量化流域尺度的地層下陷問題。本研究利用大量鑽孔資料，採用一維連續馬可夫鏈方法、spMC套件和地質統計方法來產生水文地質模型。使用地下水模擬系統(groundwater modeling system, GMS)軟體搭配TPROGS、MODFLOW和SUB套件，可模擬暫態地下水流並模擬地層下陷。透過蒙地卡羅模擬法進行模擬與評估，以確定變量的結果和不確定性。
在局部尺度中，本研究結果強調了水文地質模型的不確定性對地下水流和地層下陷模擬的重要性的顯著影響。研究中也探討了邊界條件對模擬結果的影響，給定數值的邊界條件增強了不同模型中地下水流的穩定性並減少了不確定性。在盆地尺度中，研究結果強調，較高的鑽孔密度會產生更詳細和複雜的水文地質模型，從而增強地質不連續性的代表性。相反，較低的鑽井密度會導致較連續和均勻的水文地質模型，可能會過度簡化水文地質的複雜性，但足以進行較大尺度的水資源評估。本研究強調了鑽孔密度在降低模型不確定性和提高水文地質模型可靠性方面的關鍵作用。 此外，本研究亦發展了地下水流和地層下陷的非耦合模型，透過率定和驗證步驟，使此模型的結果顯示出很高的可信度。模擬顯示，淺層地下水抽取顯著影響深部的含水層，特別是在台灣高鐵沿線，顯著導致地層下陷。 結果顯示，淺層的抽水活動引致6%至35%深層壓縮，其量值取決於局部的水文地質特徵和抽水行為。
本研究深入探討了影響地下水流和地層下陷的行為，強調了水文地質模型不確定性、鑽孔密度和地下水抽水在水文地質學研究中的重要性。這些發現對於面臨嚴重地下水缺乏和人為影響地區的地下水管理、環境影響和地層下陷評估具有重要意義。未來的研究應整合其他自然因素，例如水文循環的變化，以進一步改善對地層下陷和地下水資源管理的理解和緩解策略。;In the complex aquifer systems of the Choushui River Alluvial Fan, Taiwan, effective groundwater management and land subsidence mitigation are crucial in the fields of geohydraulic, engineering geology, and environmental science. This study aims to clarify the impact of heterogeneous hydrogeological models on geohydrology, spanning from data to modeling. Our objectives address three primary challenges: (1) quantifying the influence of uncertainty of heterogeneous hydrogeological model on local-scale simulations of groundwater flow and land subsidence; (2) assessing the impact of borehole density on the construction and effectiveness of three-dimensional heterogeneous hydrogeological models in the basin scale; and (3) developing an uncoupled-model for groundwater and compaction that quantifies land subsidence at a basin scale. This study utilizes extensive borehole data, employing the one-dimensional (1D) continuous-lag Markov chain, the spMC package, and geostatistical methods to generate realistic hydrogeological model realizations. The Groundwater Modeling System (GMS) software, integrated with the TPROGS, MODFLOW, and SUB packages, enables simulations of transient groundwater flows and models land subsidence. These models are evaluated through Monte Carlo simulations to ascertain the variability and uncertainty of the results.
In the local scale, our findings highlight the significant impact of hydrogeological model uncertainty on the reliability of simulations for groundwater flow and land subsidence. We also explore the effects of boundary conditions on simulation outcomes, noting that assumed boundary conditions enhance the stability of groundwater flow across different models and reduce uncertainty. At the basin scale, findings highlight that higher borehole densities produce more detailed and complex hydrogeological models, enhancing the representation of geological discontinuities. Conversely, lower densities result in more continuous and uniform patterns, potentially oversimplifying subsurface complexities, yet sufficient for expansive assessments. This study emphasizes the critical role of borehole density in reducing model uncertainty and enhancing the reliability of hydrogeological model. Furthermore, an uncoupled-model for groundwater and compaction was well developed with the calibration and verification. Results from the model have demonstrated high credibility. The simulations reveal that shallow groundwater pumping significantly impacts deeper hydrogeological layer, notably along the Taiwan High-Speed Rail, contributing markedly to land subsidence. It shows that pumping activities within shallow layers account for 6% to 35% of deep compression, with variations dependent on specific hydrogeological characteristics and pumping behaviors.
In conclusion, this research provides deep insights into the dynamics affecting groundwater flow and land subsidence, emphasizing the essential roles of uncertainty of hydrogeological model, borehole density, and groundwater pumping practices in the geohydrology study. The implications of these findings are significant for groundwater management, environmental assessments, and land subsidence evaluations in regions facing severe hydrogeological and anthropogenic changes. Future research should integrate additional natural factors, such as variations in the hydrological cycle, to further refine the understanding and mitigation strategies for land subsidence and groundwater resource management. |
