Premium
Simulating field‐scale moisture flow using a combined power‐averaging and tensorial connectivity‐tortuosity approach
Author(s) -
Zhang Z. Fred,
Khaleel Raziuddin
Publication year - 2010
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/2009wr008595
Subject(s) - tortuosity , hydraulic conductivity , anisotropy , vadose zone , isotropy , geology , plume , moisture , flow (mathematics) , soil science , water content , mechanics , geotechnical engineering , meteorology , physics , soil water , porosity , optics
Various methods have been developed over the past 3 decades to estimate effective unsaturated hydraulic properties. We developed an alternative practical approach to estimate three‐dimensional effective unsaturated hydraulic conductivity via a combined power‐averaging and tensorial connectivity‐tortuosity (PA‐TCT) model. An application of the PA‐TCT model to data collected at a field injection site suggests that the model provides a reasonable framework for upscaling core‐scale measurements as well as an accurate simulation of moisture flow in a heterogeneous vadose zone. The heterogeneous media at the injection site are composed of multiple geologic units, each of which is represented by an anisotropic equivalent homogeneous medium (EHM). The directional effective hydraulic conductivity for each anisotropic EHM was determined by upscaling the laboratory‐measured hydraulic properties with the combined PA‐TCT approach. A larger difference between the power values in the horizontal and vertical directions indicates a larger macroscopic anisotropy in unsaturated hydraulic conductivity. A moment analysis was used to quantify the center of mass and the spread of the injected water. Numerical simulations showed that, if the flow domain was treated as being isotropic, the vertical migration was significantly overestimated while the lateral movement was underestimated. To the contrary, if the media were treated as layered, the lateral moisture movement was considerably overestimated while the vertical movement was underestimated. However, when the flow domain was modeled as being mildly anisotropic with the PA‐TCT‐based parameters, the model could successfully predict the moisture flow and the simulated plume matched best the center of mass and the spread of the injected water of the observed moisture plume.