The Use Of Ground Penetrating Radar For Aquifer Characterization: An Example From Southwestern British Columbia

Abstract

Twelve kilometers of ground penetrating radar (GPR) data have been collected over the Brookswood aquifer in southwestern British Columbia. The data have been analyzed to assess how GPR can be used to characterize the distribution and connectivity of hydraulic units. We have used GPR to locate the aquifer/aquitard boundary at several locations in the study area. The electrical contrast between these two materials makes the aquifer/aquitard boundary an excellent target for GPR surveys. GPR was also used to reconstruct the paleo-environment of one area of the Brookswood aquifer. This was accomplished by using a modification of the concept of architectural element analysis. Radar elements were identified in the survey and were assigned sedimentary parameters using data from trenching and drilling in the area. These elements were used to develop an interpretation of the paleo-environment that provides information about the spatial distribution of hydraulic units. INTRODUCTION Hydrogeologists require quantitative data to produce a realistic model of the spatial variabilities in hydraulic properties of an aquifer. Such data can be difficult and expensive to obtain. A possible solution is to develop geophysical techniques as a means of aquifer characterization. Ground penetrating radar (GPR), a shallow geophysical technique, is well suited for this purpose as it can be used to image to a depth of up to 30m in sand and gravel environments. However, the image produced by a GPR survey does not supply hydrogeologic parameters directly. The focus of this paper is to investigate how GPR can be used for aquifer characterization. At an aquifer scale of lo’s to 100’s of meters, the most fundamental aspect of aquifer characterization is the determination of the aquifer’s hydraulic connectivity through mapping of aquifer/aquitard interfaces. GPR can be used for this purpose due to the large contrast in electrical conductivity between the sand and gravel material of an aquifer, and the clay rich material of an aquitard. The electrical conductivity of a material affects the penetration depth of radar waves, such that radar waves penetrate well through resistive material, but poorly through conductive material. Aquifers, composed of sands and gravels, are resistive, while aquitards, composed of clay rich materials, are electrically conductive. Therefore a radar survey will show good penetration in aquifer materials and very poor penetration in aquitards. By exploiting this difference in radar response, the aquifer boundaries can be mapped. At a smaller scale of centimeters to meters, the determination of the internal structure of an aquifer is also important for aquifer characterization. For example, anisotropy within the aquifer can cause significant differences in hydraulic properties and so must be identified where present. In addition, identification of sedimentary features aids in the determination of the paleo-environment that can provide important insight into the probable arcal extent and orientation of geological units. GPR can be used to image these features because of changes in their electrical properties. GPR and Sedimentology GPR has received considerable attention as a means of imaging sedimentary stratigraphy (Jo1 and Smith, 1992; Smith and Jol, 1992; Pratt and Miall, 1993; Greenhouse et al, 1987; Rea et al, 1991; Huggenberger et al, 1994). The key question that needs addressing is exactly which sedimentary aspects of the subsurface are imaged with GPR. A GPR survey, conducted by transmitting radar waves into the subsurface and recording the reflected energy, will image changes in the subsurface dielectric constant and conductivity. If these electrical properties correspond to changes in sedimentary parameters, then a GPR survey can be said to image sedimentary features. The dielectric constant and conductivity of earth materials are dependent upon composition and geometry of the solid and liquid components. Sedimentological classification is based upon five fundamental properties from which all others can be derived: grain composition, size, shapes, orientation and packing (Blatt et al, 1980). These five properties clearly are related to the composition and geometry of the solid component of a system. It is therefore reasonable to assume that a change in sedimentological properties at some boundary will cause a change in electrical properties. If the resulting change in electrical properties is large enough, then the sedimentary boundary will be imaged in a radar survey. The complicating issue is the liquid, usually water, component which does not play a role in sedimentary