Determination of the influence of soil parameters and sample density on ground-penetrating radar: a case study of a Roman picket in Lower Bavaria

Ground-penetrating radar (GPR) surveys are very much influenced by the site-specific physical properties, soil parameters and the sample density of the radar pulses. According to the electromagnetic theory, the transmission and reflection of electromagnetic waves is dominated by the conductivity, dielectric permittivity and the soil moisture. Therefore, it is advisable to measure these parameters in advance of a GPR survey. Commonly, this is not possible directly with the standard common-offset GPR arrays. To overcome this problem, we adapted and applied a direct method called time-domain reflectometry to a selected case study of the Roman picket Hienheim at the Raetischer Limes in Lower Bavaria. Furthermore, we present some important results relating to the profile spacing and orientation of GPR surveys and the site-specific soil parameters. The location and the identification of the archaeological features were successful by all geophysical methods. However, the application of all these geophysical methods like magnetometry, resistivity and GPR in addition with aerial photography and Airborne Laserscanning exemplifies the advantages of an integrated survey to achieve a comprehensive result of an archaeological site. The Roman picket of Hienheim shows up differently in all results, as each one traces the archaeological remains according to a specific physical parameter. Resistivity and radar are best to resolve buried stone structures like the walls of the picket or stone-filled ditches. In contrary, magnetometry benefits of the magnetic contrast between the limestone walls and the surrounding soil.