本研究自一個既有人造三維地質模型(Synthetic model)產生大量地質鑽孔資料，使用GMS(Groundwater Modeling System)軟體建立地質概念模型，並作為真實地質模型，根據(1)有無使用區域地質知識修正地質概念模型、(2)是否給予鑽孔正確的沉積層序(Horizon ID)，以及(3)不同的鑽孔密度提供之地層層界，建立不同的地質概念模型，並依各地層的異向性設定水力傳導係數，據以建立不同的地下水流模型，以MODFLOW模式搭配MT3DMS模式進行1,1-二氯乙烯與三氯乙烯地下水污染物傳輸模擬，以進一步探討前述三項因素對地質概念模型建置以及對污染傳輸模擬之影響。 研究成果顯示，即使有387個鑽孔，仍無法使用GMS軟體內外插出良好的層界不整合面，而使用GMS軟體自動產生鑽孔層序的方法，比使用人為給定正確沉積層序的方法，需要更大量的鑽孔才能內外插出合理的地質概念模型，因此若能正確判斷鑽孔層序，將可使用更少的鑽孔建立合理的地質概念模型。此外，本研究透過所建立的地質概念模型，比對研究區既有監測井開篩段地層，並與觀測水位與降雨量綜合研判，剔除非開篩在同一含水層之監測井，以剩下的29口監測井觀測水位作為地下水流模型之率定標的。不同地下水流模型，率定後RMSE(Root Mean Square Error)差距在1.4公尺以內，然而各模型地下水流場並不相同，且固定深度的污染源，在不同模型中所在的地層亦不相同，因此各模型污染物傳輸模擬結果具有明顯差異，顯示水文地質概念模型對於污染傳輸模擬具有重要影響。 ;In this study, a three-dimensional synthetic geological model was used to generate a lot of borehole drillings data. The GMS (groundwater modeling system) software was used to establish a geological models. The discussion of differences between models was made whether the different density of borehole drillings, the knowledge of regional geology and given the correct sedimentary sequence (Horizon ID) of the layers of borehole drillings was used or not. The numerical models of groundwater flow models were construct, with setting the hydraulic conductivities according to the anisotropy of each layer and by using MODFLOW code, and MT3DMS code was used to simulate the transport of 1,1-Dichloroethylene and Trichloroethylene. The results show that a well unconformity cannot be extrapolated and interpolated even if there were 387 borehole drillings data (approximately 5 hectares 1 borehole). Using GMS software to automatically generate sedimentary sequence to construct a reasonable geological conceptual model required more borehole drillings data than using the correct sedimentary sequence. Therefore, if the sedimentary sequence of the borehole drillings data can be correctly judged, it will be possible to use fewer borehole drillings data to construct a reasonable geological conceptual model. In addition, through the constructed geological conceptual model, this study compares the layers of the screen section of the monitoring wells in the study area, and comprehensively judges the observation data and rainfall data to excluded the monitoring wells that are not screen in the same aquifer system. Groundwater flow models were calibrated by using the observation data from in 29 groundwater monitoring wells and the RMSE (Root Mean Square Error) in each model is within 1.4 meters, the groundwater flow direction is different. The layer of contaminant source was also different in different geological models. Therefore, the simulation results of contaminant transport have obvious differences. The study results illustrated that the hydraulic geological model has an important influence on the simulation of groundwater contaminant transport.
