Characterization of Karst Features Using Electromagnetics and Microgravity: A Strategic Approach

The flat-lying limestones of western Texas are naturally jointed with preferential dissolution occurring along joints and bedding planes. This area has some of the largest air-filled open caves in West Texas and large paleokarst collapse features have been identified throughout a wide region. Aerial photos indicate the presence of extensive joints, lineaments and paleokarst. Site characterization to detect and map such subsurface features using borings alone is inadequate to produce a reasonable level of spatial sampling. This paper addresses the application of surface geophysical techniques, and an assessment of the spatial sampling and instrument sensitivities necessary to define the karst features of interest in this geologic setting. Electromagnetic measurements using a Geonics EM34 were selected because the measurements provide an excellent means of locating dissolution-enlarged joints. Microgravity was selected because it is the only surface geophysical method that will provide the location of karst feature regardless of their shape or fill material as long as there is a sufficient density contrast. While both methods have limitations, as do all methods of site characterization, the combination of gravity and EM measurements are complementary in this application. A known cave system was used to establish the spatial sampling criteria for the detection and characterization of such features. Field tests were then run over two known karst sites; a paleokarst collapse; and a localized doline to provide anomaly signatures in this geologic setting. It is clear that the EM and the gravity techniques are appropriate for detecting and characterizing karst features in this geologic setting. However, one of the key issues in planning and carrying out a geophysical survey, is developing a spatial sampling criteria. This criteria should be based upon an understanding of project objectives, a conceptual geologic model of site conditions and, if possible, existing data from the area of interest.