Book excerpt: DOE faces the remediation of numerous contaminated sites, such as those at Hanford, INEEL, LLNL, and LBNL, where organic and/or radioactive wastes were intentionally or accidentally released to the vadose zone from surface spills, underground tanks, cribs, shallow ponds, and deep wells. Migration of these contaminants through the vadose zone has lead to the contamination of or threatens to contaminate underlying groundwater. A key issue in choosing a corrective action plan to clean up contaminated sites is to determine the location, total mass, mobility and travel time to receptors for contaminants moving in the vadose zone. These problems are difficult to solve in a technically defensible and accurate manner because contaminants travel downward intermittently through narrow pathways driven by variations in environmental conditions. These preferential pathways can be difficult to find and predict. The primary objective of this project is to determine if and when dynamical chaos theory can be used to investigate infiltration of fluid and contaminant transport in heterogeneous soils and fractured rocks. The objective of this project is being achieved through the following Activities (1) Evaluation of chaotic behavior of flow in laboratory and field experiments using methods from non-linear dynamics; (2) Evaluation of the impact these dynamics may have on contaminant transport through heterogeneous fractured rocks and soils, and how it can be used to guide remediation efforts; (3) Development of a conceptual model and mathematical and numerical algorithms for flow and transport, which incorporate both: (a) the spatial variability of heterogeneous porous and fractured media, and (b) the description of the temporal dynamics of flow and transport, which may be chaotic; and (4) Development of appropriate experimental field and laboratory techniques needed to detect diagnostic parameters for chaotic behavior of flow. This approach is based on the assumption that spatial heterogeneity and flow phenomena are affected by non-linear dynamics, and in particular, chaotic processes. The scientific and practical value of this approach is that the authors can predict the range within which the parameters of flow and transport change with time in order to design and manage the remediation, even when the authors can not predict the behavior at any point or time. This progress report summarizes work after 1.5 years of a 3-year project.