由於近年來的環境問題，例如各種尺度下的揮發性油污染清除(VOCs)或放射性廢棄物及二氧化碳(CO2)封存等，使得氣體在非飽和孔隙介質的傳輸行為，受到科學與工程領域極大的重視。這些問題無論尺度大小，都需足夠的地層參數如空氣滲透係數及孔隙率等分布，作為預測與工程設計之依據。由過去的研究及現地經驗顯示，精確地描述土壤異質性及彌補資料分析能力上的不足，是提高預測準確度及工程效益的首要問題；亦即，透過有限的量測數據，考慮天然土壤層的異質性，界定資料點不足或參數異質性所造成的推估不確定性，整合不同性質數據提高觀測數據之效益，以及規劃高效益量測點位置及資料型態以取得最佳的地層參數資訊。根據以上的目標，本研究主要目的乃透過數值及實驗方式建立、測試及驗證一個以氣壓剖面掃描概念為基礎的序率反推估模式，此模式考慮空氣於非飽和土壤之壓縮性及流動特性，整合一系列的注/ 抽氣觀測資料進行非飽和孔隙介質參數反推估。本計畫將延續前一年之研究成果建構三維參數反推估模式及相應之砂箱試驗，利用實驗及數值模式量化量測資料在空間與不同維度之間的關係，以及推估量測數據之不確定性提供最佳量測點分析；計畫之第二年將模式整合訊息傳送介面(MPI)成為平行計算模式，透過叢集系統的計算提昇計算效能及擴大模擬尺度。本研究之成果除發展一新的反推估模式外，將有助於深入瞭解氣體在異質性非飽和孔隙介質的傳輸特性，並量化異質性與資料點不足在預測污染傳輸時所引入之誤差。 ; The behaviors of gas phase transport in unsaturated porous media have received increasing attentions recently due to the environmental issues such as removing volatile organic compounds (VOCs) and evaluating potential sites for storing radioactive waste and carbon dioxide (CO2) on different scales. Sufficient observations of permeability and porosity in such media are usually in great demand for better predictions of gas transport and optimal design of remediation procedures at sites of interests. Over the years, theories and in situ studies have revealed that accurate predictions of transport behaviors in unsaturated media highly depend on how and to what degree we predict the media heterogeneity, the based on available measurements, i.e., we aim to use limited data to involve the natural heterogeneity with higher resolution, to identify uncertainties due to the heterogeneity and limited observations, to incorporate different sources of data to improve estimations, and to provide optimal locations or data types to efficiently represent the system responses. Based on these objectives, the purpose of this project is to develop, test, and validate a stochastic inverse model based on the concept of pneumatic tomography surveys. The developed model accounts for compressibility and density of air and estimates the hydrgeologic parameters in unsaturated porous media, using air pressure measurements from sequential cross-hole pneumatic pumping or injection tests. This project will extend preview study to further improve our model to be a three-dimensional one and verify it with three-dimensional sandbox experiments. In the second year we will incorporate our three-dimensional mode with message passing interface (MPI) and excute the model on a cluster system to improve the calculation efficiency and extend the modeling scales. The expected contributions of this project will be developing a new gaseous-based inverse model and providing in depth analyses of effects on heterogeneity and limited observations to predict contaminant migration in unsaturated porous media. ; 研究期間 9708 ~ 9807
