Identifying Unsaturated Hydraulic Parameters Using an Integrated Data Fusion Approach on Cross-Borehole Geophysical Data

238 C geophysical methods are increasingly used to estimate hydrogeologic properties of the unsaturated zone. Th ese methods provide high-resolution in situ information for large support volumes of the subsurface with minimal intrusion and thereby have obvious advantages over other more traditional methods such as neutron probe and time domain refl ectrometry. In particular, cross-borehole electrical resistance tomography (ERT) and ground penetrating radar (GPR) have become very popular in the past decades. Applications of these methods range from characterization of permeable pathways in fractured basalt (Hubbard et al., 1997) and monitoring preferential fl ow during early stages of snowmelt (French et al., 2002) to following the eff ect of diff erent remediation processes (Daily et al., 1995; LaBrecque et al., 1996). Th e moisture content estimates derived from GPR have been shown to have a high reproducibility and accuracy (Parkin et al., 2000; Ferré et al., 2003). As a result, water front movement caused by natural infi ltration can be monitored successfully despite just slight variations in moisture content (Binley et al., 2002b). Cross-borehole GPR and ERT are especially suitable for monitoring relative changes in soil moisture and electrical conductivity, since local variations in the petrophysical relationships are, in this way, reduced or completely avoided (Binley et al., 2002a). Th erefore many studies have been performed using the two methods to investigate the fl ow dynamics of the unsaturated zone resulting from forced loading to the system (e.g., Daily et al., 1992; Binley et al., 2001; Alumbaugh et al., 2002). Water fl ow patterns have been visualized and attempts have also been made to quantify important fl ow and transport parameters, especially the eff ective hydraulic conductivity (Ks) (Binley et al., 2002a; Cassiani et al., 2004: Kowalsky et al., 2004, 2005; Rucker and Ferré, 2004; Cassiani and Binley, 2005; Cassiani et al., 2006). Estimation of subsurface hydraulic parameters can be performed by combining moisture content data from cross-borehole GPR with collected bulk electrical resistivity data obtained from crossborehole ERT data (Winship et al., 2006). Looms et al. (2008) applied the two methods simultaneously for determination of fl uid conductivity and thereby tracer concentration, which was subsequently used for moment analysis of the evolving plume. Th e tomographic images resulting from cross-borehole geophysical surveys are, however, subject to several uncertainties and artifacts caused by the inversion techniques used to invert the Identifying Unsaturated Hydraulic Parameters Using an Integrated Data Fusion Approach on Cross-Borehole Geophysical Data