This paper describes and demonstrates two methods of providing a priori information to the surface‐based time‐lapse three‐dimensional electrical resistivity tomography (ERT) problem for monitoring stage‐driven or tide‐driven surface water intrusion into aquifers. First, a mesh boundary is implemented that conforms to the known location of the water table through time, thereby enabling the inversion to place a sharp bulk conductivity contrast at that boundary without penalty. Second, a nonlinear inequality constraint is used to allow only positive or negative transient changes in EC to occur within the saturated zone, dependent on the relative contrast in fluid electrical conductivity between surface water and groundwater. A 3‐D field experiment demonstrates that time‐lapse imaging results using traditional smoothness constraints are unable to delineate river water intrusion. The water table and inequality constraints provide the inversion with the additional information necessary to resolve the spatial extent of river water intrusion through time.

Semiarid ecosystems play a critical role in determining the interannual variability of the global terrestrial carbon sink. Water availability is a critical driver of productivity in semiarid ecosystems, which often alternate between carbon sink/source functioning during wet/dry years. In this study, we investigate how groundwater availability resulting from groundwater‐river water exchange influences net ecosystem exchange of CO2 (NEE), evapotranspiration (ET), and the surface energy balance at two semiarid ecosystems along the Columbia River in central Washington, USA. We examined 1 year of eddy covariance measurements from an upland sagebrush ecosystem primarily fed by rainfall without groundwater access and a riparian grassland ecosystem with groundwater access during the dry season due to lateral groundwater‐river water exchange. The two sites had distinct seasonal patterns of NEE and ET, driven by differences in water availability. While NEE at the upland sagebrush site was strongly constrained by water availability during the dry months, access to groundwater allowed the riparian site to maintain high NEE magnitude and ET during the same dry months. The riparian site had larger annual gross primary productivity than the upland site (612 vs. 424 gC/m2), which was offset by higher ecosystem respiration (558 vs. 363 gC/m2). Thus, the magnitude of the annual NEE at the upland site was larger than that at the riparian site (−62 vs. −54 gC/m2). Our results demonstrate that groundwater access determined by connectivity between groundwater and surface water can be a critical driver of carbon uptake and ET in semiarid ecosystems.

Heterogeneity of hydraulic properties in aquifers may lead to contaminants residing in lower‐permeability zones where it is difficult to deliver remediation amendments using conventional injection processes. The focus of this study is to examine use of a shear‐thinning fluid (STF) to improve the uniformity of remedial amendment distribution within a heterogeneous aquifer. Previous studies have demonstrated the significant potential of STFs for improving remedial amendment delivery in heterogeneous aquifers, but quantitative evaluation of these improvements from field applications is lacking. A field‐scale test was conducted that compares data from successive injection of a tracer in water followed by injection of a tracer in an STF to evaluate the impact of the STF on tracer distribution uniformity in the presence of permeability contrasts within the targeted injection zone. Data from tracer breakthrough at multiple depth‐discrete monitoring intervals and electrical resistivity tomography (ERT) showed that inclusion of STF in the injection solution improved the distribution of the injected fluid within the targeted treatment zone. One improvement was a reduction in the movement of injected fluids through high‐permeability pathways, as evidenced by slower breakthrough of tracer at monitoring locations where breakthrough in baseline tracer‐only injection data was faster. In addition, STF‐amended injection solutions arrived faster and to a greater extent in monitoring locations within low‐permeability zones. ERT data showed that the STF injection covered a higher percentage of a two‐dimensional cross section within the injection interval between the injection well and a monitoring well about 3 m away.

Current plans for treatment and disposal of immobilized low-activity waste (ILAW) from Hanford's underground waste storage tanks include vitrification and storage of the glass waste form in a near-surface disposal facility. This Integrated Disposal Facility (IDF) is located in the 200 East Area of the Hanford Central Plateau. Performance assessment (PA) of the IDF requires numerical modeling of subsurface flow and reactive transport processes over very long periods (thousands of years). The models used to predict facility performance require parameters describing various physical, hydraulic, and transport properties. This report provides updated estimates of physical, hydraulic, and transport properties and parameters for both near-and far-field materials, intended for use in future IDF PA modeling efforts. Previous work on physical and hydraulic property characterization for earlier IDF PA analyses is reviewed and summarized. For near-field materials, portions of this document and parameter estimates are taken from an earlier data package. For far-field materials, a critical review is provided of methodologies used in previous data packages. Alternative methods are described and associated parameters are provided. For far-field materials, consisting of both sand-and gravel-dominated facies underlying the IDF, a particular model has been used in previous PA modeling efforts to represent the saturation-dependent anisotropy of unsaturated hydraulic conductivity. We recommend that this model be replaced with a more recent and general tensorial pore-connectivity-tortuosity (TCT) model for saturation-dependent anisotropy. Simulation results from both the TCT and the earlier anisotropy model have been compared with observed data from a controlled vadose zone field injection experiment performed just south of the 200 East Area. The TCT model was shown to predict observed flow behavior at this site as well as or better than the model used in previous PA efforts, and with many fewer added model parameters (one versus eight). Recommended parameter estimates for the TCT model are presented. Previous estimates of dispersivities for vadose zone sediments were based on stochastic theory developed for saturated aquifer materials. An extensive literature review is presented that suggests these estimates may not be appropriate for unsaturated conditions. An alternative approach based on more fundamental physical property information is described and updated parameter estimates are presented.

No abstract