3D inversion of ERT data on an archaeological site using GPR reflection and 3D inverted magnetic data as a priori information

Non-destructive geophysical prospecting methods are increasingly used for the investigation of archaeological sites, where a detailed physical and geometrical reconstruction of structures is required prior to any excavation work. In 1996, at the Archaeological Test Site of the Sabine Necropolis at Research Area of National Research Council of Rome (Montelibretti, Italy), a small area had been selected to test the integration of magnetic, ground-penetrating radar, and electrical resistivity tomography techniques. The magnetic data were acquired as the vertical gradient of the total magnetic field and were inverted by minimizing an L-2 norm. The ground-penetrating radar data were acquired as singleoffset measurements on a regular grid and processed to obtain time slices of reflection amplitude. The electrical resistivity tomography data were acquired with dipole–dipole array along parallel profiles in both east–west and north–south directions. In this work, we reprocessed the data collected in 1996 performing a three-dimensional inversion of electrical resistivity tomography data using ground-penetrating radar data and three-dimensional inversion results of the magnetic data. We inverted electrical resistivity tomography data according to L2 and L1 norm minimization. Both the mean apparent resistivity and different resistivity distributions derived from processed ground-penetrating radar and three-dimensional inversion of the vertical gradient of the total magnetic field data were used as a starting model. The results were then compared in the volume under the surveyed area: merits and shortcomings of the different techniques are pointed out and analyzed. From the results of the different applied techniques, it comes that both electrical resistivity tomography and vertical gradient of the total magnetic field locate the searched tomb as a volume, whereas ground-penetrating radar mainly identified the ceiling of the tomb. Vertical gradient of the total magnetic field data inversion obtained from minimization of the L2 norm properly locates both the chamber and the corridor but produces smeared overestimated volumes. Three-dimensional inversions of electrical resistivity tomography data with a priori information give an accurate picture of the chamber and delineates the corridor, even if the resistivity values are underestimated if compared with the expected resistivity of void. Results are compared with the archaeological findings as the area was unearthed and verified in 2001.