ADVANCING SUBSURFACE IMAGING TECHNOLOGY: EVALUATING GPR ACCURACY AGAINST LIDAR MODELS AT KHIRBAT ATARUZ, JORDAN

The use of high-resolution geophysical imaging in detecting and mapping cavernous spaces is an emerging technology. We present preliminary results integrating Ground Penetrating Radar (GPR) data with 3D Light Detection and Ranging (LiDAR) images captured of cave systems at the celebrated 3,000-years old archaeological site of Khirbat Ataruz in the Madaba Governorate, Jordan, acquired in June and July of 2024. The surface geological stratigraphy is Late Cretaceous-age limestone interspersed with occasional chert layers. Owing to the dry climate, little to no soil has developed or been preserved where GPR surveys were acquired. The predominantly limestone bedrock has extensive development of caverns and small sinkholes or dissolution pits that surround and underlie Khirbat Ataruz. Cave habitation by previous populations is evinced by man-made materials, including walls, arches, and ceiling ducts, which complexity requires sophisticated imaging technology. Using Polycam LiDAR imagery, 3D models and top-down views (blueprints) of various caves were digitally built, while 2D profiles of these same caves were then created using GPR above each cave. The site is well-suited for GPR surveying due to the summertime absence of moisture, low vegetation cover, lack of clay and salinity, and the availability of smooth, flat ground. The GPR profiles are analyzed alongside the LiDAR models to assess how well GPR identifies and characterizes underground voids. The GPR surveys used a GSSI (Geophysical Survey Systems, Inc.) 400-MHz bistatic antenna in continuous mode. The listening time was 50 ns (equivalent to 1.94 m depth, assuming a dielectric constant = 15 or velocity ~ 0.077 m/ns) with 1024 samples per scan (~0.05 ns sample interval) and 36 scans per meter. Field filters were set at 100-800 MHz bandwidth. This research will allow geologists and archaeologists to have a greater understanding of how complex subsurface void structures can be represented by geophysical imaging. By using caves with known spatial data, provided by Polycam LiDAR imaging, this study highlights the potential of GPR subsurface mapping accuracy.