Magnetic and radar surveys at Locri Epizephyrii: A comparison between expectations from geophysical prospecting and actual archaeological findings

In recent years, geophysical methods have been increasingly applied as a preliminary mapping tool to guide archaeological excavations. Despite the reported growing case histories of geophysical prospections for archaeological purposes, direct comparison between expectations arising from the geophysical results and actual findings is not always systematically performed. A critical comparison between pre-excavation geophysical-guided hypotheses and post-excavation archeological evidences is proposed in this work. A test site within the archaeological area of Locri Epizephyrii (Calabria, southern Italy) was chosen for this purpose. An unsurveyed rectangular area (31 × 26 m) was investigated with high-density ground-penetrating radar (GPR) and magnetic profiles. Several anomalous alignments, both compatible and oblique with respect to the orientation of the Greek-Roman city plan, were preliminarly observed in the geophysical results. The use of two different techniques allowed for comparison of anomalous areas, enhancing the likelihood of finding features of significance. Two archaeological soundings were later carried out in the areas showing the most peculiar geophysical anomalies. Several structures revealing the same orientation of the ancient city plan and belonging to at least two different building phases were unearthed from a depth of 15–25 cm below the ground surface. A systematic comparison between geophysical and archeological results was then carried out. In general, walls showing different construction typologies (opus testaceum, opus incertum or opus caementicium) were found to generate similar radar anomalies. Artifacts made with materials similar to the background sandy alluvial deposits were not identified by both geophysical techniques, as in the case of an unearthed channel bank, made of local sandstone blocks. GPR was globally observed to detect buried structures better than the magnetic method, probably due to background geological variables linked to the presence of Fe-rich minerals within the background sediments, generating noisier and scattered magnetic gradient maps.