HR: 1340h
AN: H33F-0530    [Abstracts]
TI: Streamline based analysis of non-Fickian dispersion in porous media with multi-scale heterogeneity
AU: * INOUE, J
EM: inoue@ohriki.t.u-tokyo.ac.jp
AF: The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo, 113-8656 Japan
AU: MOTOSHIMA, T
EM: mtstky00@pub.taisei.co.jp
AF: Taisei Co.Ltd., 1-25-1 Nishi-Shunjuku, Shinjuku, Tokyo, 163-0613 Japan
AU: Chun, P
EM: chun@ohriki.t.u-tokyo.ac.jp
AF: The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo, 113-8656 Japan
AB: A numerical method to model contaminant transport in heterogeneous geological formations is developed. The present model is based on the framework that takes into account of two different spatial scales of uncertainty, microscopic level and macroscopic level. The microscopic heterogeneity caused by the existence of a soil particle and a crack, which is considered smaller than the scale of the measurement resolution hence unresolved in the real application, is treated by a stochastic method, while the microscopic heterogeneity, which can be treated as a large scale trend and well defined, is solved deterministically by the streamline method. The stochastic method developed is based on the continuous time random walk (CTRW) formalism, which is numerically solved along a streamline. The application of the streamline method to the integration of CTRW in the space domain is based on the fact that the macro-dispersion tensor is usually observed highly anisotropic and that the transverse dispersivity can be negligible compared to the longitudal dispersivity. Since the integration in the space domain is assumed 1-D in the streamline method, the numerical integration can be simplified by applying the Laplace transformation in the time domain. In obtaining a streamline numerically, the mesh dependency observed in the classical finite element approach especially in the case of the unstructured grids has to be eliminated. In the present method, the element free approach is adopted. The accuracy of streamlines obtained by the element free method is evaluated through the comparison with the analytical solution for the repeated spot pattern. The calculations are compared with the experimental results, in which the statistical distribution of the conductivity is well controlled. From the comparison with the experimental result, not only the validity of the present model is discussed but also the effect of the anomalous transport characteristics to the macro-dispersion is clarified.
DE: 1829 Groundwater hydrology
DE: 1832 Groundwater transport
DE: 1869 Stochastic processes
DE: 1875 Unsaturated zone
SC: Hydrology [H]
MN: 2004 AGU Fall Meeting