HR: 1340h
AN: H23A-1006 [Abstracts]
TI: Use of high-resolution geophysical data to characterize the porosity distribution in heterogeneous aquifers: Influence of inversion and data integration method on hydrological predictions
AU: * Dafflon, B
EM: baptiste.dafflon@unil.ch
AF: Institute of Geophysics, University of Lausanne, Lausanne, 1015, Switzerland
AU: Irving, J
EM: james.irving@unil.ch
AF: Institute of Geophysics, University of Lausanne, Lausanne, 1015, Switzerland
AU: Holliger, K
EM: klaus.holliger@unil.ch
AF: Institute of Geophysics, University of Lausanne, Lausanne, 1015, Switzerland
AB:
Knowledge of the detailed distribution of hydrological parameters in heterogeneous aquifers is a key prerequisite
for accurate simulation of groundwater flow and contaminant transport. The integration of high-resolution
geophysical data into the subsurface characterization problem has been shown in many cases to significantly
improve our knowledge of hydrological parameters by providing information at spatial scales that are unattainable
using conventional measurement techniques. However, the hydrological significance of many of the choices
made during the complex process of geophysical data integration (e.g., data processing and inversion
techniques, integration/simulation methodology) has not been fully evaluated. In addition, evaluation of how
much benefit is brought by the geophysical data for various types of hydrological models has not been thoroughly
investigated. Understanding these issues is critical to making wise decisions about how to use the geophysical
data.
Here, we examine some of these issues for the case of porosity characterization in saturated heterogeneous
aquifers using crosshole ground-penetrating radar (GPR) and borehole porosity log data. To begin, we generate
a number of different porosity fields that exhibit varying degrees of continuity and structural complexity. Next, we
simulate the collection of crosshole GPR data between several boreholes in these fields, and the collection of
porosity log data at the borehole locations. The synthetic GPR data are then tomographically inverted for the
spatial distribution of electromagnetic wave velocity using a variety of inversion techniques. Together with the
synthetic porosity logs, the resulting tomographic images are used to reconstruct the porosity field through Monte-
Carlo conditional simulations. To accomplish this, we again use a variety of methods, including sequential
simulation and simulated annealing. The resulting realizations of porosity, obtained using the various
combinations of inversion and data integration techniques, are then used as basis information to infer the
hydraulic conductivity field and perform groundwater flow and contaminant transport simulations. This allows us
to assess the hydrological significance of our choice of geophysical inversion and data integration methods.
DE: 1829 Groundwater hydrology
DE: 1835 Hydrogeophysics
DE: 3252 Spatial analysis (0500)
DE: 3275 Uncertainty quantification (1873)
DE: 6982 Tomography and imaging (7270, 8180)
SC: Hydrology [H]
MN: 2007 Fall Meeting