H32B-01
Validation of Rivulet Approach to Infiltration in Sand Boxes Using Neutron- Radiography
Rivulets (i.e., tiny water streaks) are considered the basic units of preferential infiltration. Rivulet approach is based on the hypothesis that flow shows continuous features. Rivulet approach was applied to simulate flow process. We used Neutron Radiography (NR) as a tool to monitor the infiltration-drainage process in sand boxes. Through a series of image analysis procedures, continuity/discontinuity of flow and saturation were analyzed. Sand of three different grain size (>0.2mm, >0.5mm and >1.0mm) was packed in aluminium sample holders (40x20x0.5-1.0cm). Irrigation experiments and series of radiograms were carried out in air dried samples and in samples which were equilibrated with water and deuterium prior to the experiments. We report soil moisture variations during and after infiltration, velocity of wetting fronts, the impact of the rate of infiltration on type of flow and exchange between deuterium and water. Questions whether wetting front velocities in the sand boxes, minimum water contents and the geometry of flow represent features of preferential flow will be discussed.
H32B-02
Vadose Zone Monitoring System as a Tool for Groundwater Protection
Subsurface monitoring for groundwater protection from pollution hazards has traditionally been based on culling information from the groundwater. This information is usually retrieved from boreholes penetrating the saturated section of the groundwater. Accordingly, the entire path and fate of pollutants transported from land surface through the vadose zone to the groundwater is evaluated from the chemical and physical state of the water which has been sampled from a well. That monitoring procedure is well founded in both scientific studies and through legislative acts which enforce groundwater monitoring for potential sources of pollution. However, this creates a paradox since, by definition, identification of pollution in groundwater means that the groundwater is already polluted. Moreover, since vertical transport in the vadose zone and lateral flow in the groundwater are very slow processes, pollution identification in a well may take years or decades. As a result, the total mass of pollutant that has penetrated the subsurface may be extremely high by the time it has been identified. Finally, pollution identification in a well usually reveals only the edges of a much larger pollutant plume. Accordingly, identification of pollution in the vadose zone right under the pollution source, long before it shows up in the groundwater, should be the key to groundwater protection. The need for real-time information on the quality of percolating water led to the development of a new vadose- zone monitoring system. The new monitoring system is designed to provide continuous measurements of the soil water content and water potential, while allowing pore-water sampling all along the vadose-zone cross section. The installation technique allows monitoring of the vadose-zone cross section under relatively undisturbed soil conditions. The new monitoring system is comprised of special flexible TDR (FTDR) probes, assembled with special vadose-zone sampling ports (VSPs) that function either as deep vadose-zone tensiometers or pore-water sampling devices. The system is adapted for installation in small-diameter boreholes allowing multiple measurements of the soil's hydraulic properties all along the vadose-zone cross section, from land surface to groundwater. This monitoring system has been implemented in several studies on water infiltration and groundwater recharge in different climatic and lithological setups.
H32B-03 INVITED
A Conceptual Model of Coupled Biogeochemical and Hydrogeological Processes Affected by In Situ Cr(VI) Bioreduction in Groundwater at Hanford 100H Site
The overall objective of this presentation is to demonstrate a conceptual multiscale, multidomain model of coupling of biogeochemical and hydrogeological processes during bioremediation of Cr(VI) contaminated groundwater at Hanford 100H site. A slow release polylactate, Hydrogen Release Compound (HRCTM), was injected in Hanford sediments to stimulate immobilization of Cr(VI). The HRC injection induced a 2-order-of- magnitude increase in biomass and the onset of reducing biogeochemical conditions [e.g., redox potential decreased from +240 to -130 mV and dissolved oxygen (DO) was completely removed]. A three-well system, comprised of an injection well and upgradient and downgradient monitoring wells, was used for conducting the in situ biostimulation, one regional flow (no-pumping) tracer test, and five pumping tests along with the Br-tracer injection. Field measurements were conducted using a Br ion-selective electrode and a multiparameter flow cell to collect hourly data on temperature, pH, redox potential, electrical conductivity, and DO. Groundwater sampling was conducted by pumping through specially designed borehole water samplers. Cross-borehole radar tomography and seismic measurements were carried out to assess the site background lithological heterogeneity and the migration pathways of HRC byproducts through groundwater after the HRC injection. Several alternative approaches, including conventional and fractional advective dispersion equations and geostatistical analysis, were used to characterize hydraulic and biogeochemical transport parameters. The results of a joint inversion of cross-borehole geophysical tomography and flow-rate measurements in boreholes indicate the presence of a bimodal distribution of hydraulic conductivity for Hanford sediments. The Br- concentration double-peak BTCs curves indicate that HRC injection caused an increase in the tracer travel time (mainly in the low-permeability zone) over the period of observations of about 2 years. This increase in the Br travel time could be explained by the decrease in the saturated hydraulic conductivity caused by the formation of CO2 and N2 gases, growth of biofilms, and precipitation of calcite and insoluble Cr(III) complexes. All these processes are known to cause partial blocking of flow pathways within heterogeneous media, and slowing of mobile-immobile-region mass transfer. The analysis of geophysical data was also used to delineate the temporal variations of the zone affected by byproducts of bioremediation. Our results also indicate the importance of combining in situ hydrogeological, geochemical (including stable isotope analysis), and geophysical measurements with microbiological analytical analyses of water samples and sediments. These measurements can be used for obtaining data to control and simulate both enforced biostimulation and long-term natural attenuation of metals in groundwater.
H32B-04
HYDRAULIC TOMOGRAPHY USING TEMPORAL MOMENTS OF DRAWDOWN-RECOVERY DATA: LABORATORY SANDBOX STUDY
Hydraulic tomography is a viable technology that images the hydraulic heterogeneity of the subsurface. Unlike
steady-state hydraulic tomography (SSHT), which provides estimates of hydraulic conductivity (K), transient
hydraulic tomography (THT) can provide estimates of both K and specific storage (Ss) [Liu et al., 2007]. Effective
as it may be, THT is a computationally demanding technique. To ease the computational burden, a transient
hydraulic tomography which utilizes temporal moments (THT-m) of transient drawdown-recovery data has been
developed by Zhu and Yeh [2006]. This procedure simplifies the governing equation from a diffusion equation to a
Laplace equation in the corresponding numerical analysis. However, the calculation of the temporal moments,
which involves integration in time of the drawdown-recovery data, may cause loss of information on the
parameters that are being estimated by the inverse procedure. To test this conjecture, we conduct an
investigation of the THT-m by comparing it to the THT through synthetic simulations. We then interpret a
previously conducted hydraulic tomography tests in a laboratory sandbox using the THT-m approach and
compare its performance to the results from THT previously conducted by Liu et al. [2007]. The laboratory tests
were conducted in a synthetic aquifer created with a known heterogeneity pattern with forcing functions controlled.
Our results show that the THT-m is able to estimate the K field well, but the estimation of Ss field is more difficult.
http:www.iihr.uiowa.edu/~illman
H32B-05 INVITED
Joint Inversion at the Boise Hydrogeophysical Research Site
The Boise Hydrogeophysical Research Site (BHRS) is a research wellfield or field-scale test facility developed in
a shallow, coarse, fluvial unconfined aquifer with the objectives of developing cost-effective, non-invasive
methods for quantitative characterization and imaging in heterogeneous aquifers using hydrologic and
geophysical techniques. The design of the wells and the wellfield provide for a wide range of single-well, cross-
hole, multiwell and multilevel hydrologic, geophysical, and combined hydrologic-geophysical experiments.
Recent efforts have been focused largely on: (a) establishing the 3D distributions of geologic, hydrologic, and
geophysical parameters and (b) developing subsurface measurement and imaging methods including time-
lapse tomographic imaging methods. Multiple lines of evidence from these efforts indicate that the
hydrostratigraphic framework of the BHRS is a hierarchical system with at least three scales of sedimentary
organization including layers and lenses; this framework is recognized with geologic, hydrologic, radar, seismic,
and EM methods and tracer tests. Data from these characterization efforts and experiments can be used as the
basis for joint inversion of multiple types of data to return the "known" 3D K distribution and also to improve
subsurface imaging by including prior information in addition to data collected during tomographic or time-lapse
imaging experiments. We present a joint inversion framework for determining the 3D hydraulic conductivity (K)
distribution using site K data and geophysical data. Given the heterogeneous nature of the system, K is treated
as a random field and the moment equation approach is used for the solution of the stochastic forward flow
problem. The inverse problem is posed in the Bayesian framework which allows for a straightforward inclusion
of prior information.
http:cgiss.boisestate.edu/hydrogeophysical_research_site.htm
H32B-06 INVITED
Sensitivity and Resolution of Tomographic Pumping Tests in an Alluvial Aquifer
A number of investigators have proposed hydraulic tomography, the simultaneous analysis of responses to multiple well tests or flow system stresses, as a means to obtain a higher resolution characterization of aquifer flow and transport properties than can be obtained from traditional aquifer test analysis. In this presentation I use first-order sensitivity analysis and resolution matrix analysis to quantitatively examine the resolution provided by a set of tomographic pumping tests in an alluvial aquifer. The latter technique is commonly employed in geophysical tomography and promises to be of considerable use in tomographic approaches to hydrogeologic inverse problems as well. In particular, I compare the sensitivity and resolution provided by fully transient and steady-shape approaches to the estimation of the hydraulic conductivity (K) field. The steady-shape approach takes advantage of the rapid establishment of constant gradients within the region surrounding the pumping well, comparing observed drawdown differences within this region to drawdown differences predicted by a steady- state model. The results demonstrate that either approach resolves K variations only within a certain limited distance of the pumping wells and observation points, but that, relative to the transient approach, the steady- shape approach tends to filter out the influence of unresolved property variations outside the region of investigation.