Optimization of Dominant Frequency and Bandwidth Analysis in Multi-Frequency 3D GPR Signals to Identify Contaminated Areas

Ground-penetrating radar (GPR) has been widely used in investigations of contaminated areas because of its sensitivity to variations associated with the nature of pore fluids. However, most of the studies were usually based on the visual interpretation of radargrams or on a time domain amplitude analysis. In this work, we propose a methodology that consists of analyzing the spectral content of the signal recorded in multi-frequency 3D GPR profiles. A remarkable advantage of this type of antenna is its step-frequency system, which provides a much wider emission spectrum than the one corresponding to conventional single-frequency antennas. From the data in the frequency domain, the dominant frequency and bandwidth were calculated as parameters whose variation could be related to the presence of light non-aqueous phase liquid (LNAPL) in the subsurface. By analyzing the variations of these two parameters simultaneously, we were able to delimit the contaminated zones in a case study, associating them with a significant shift of the frequency spectrum with respect to the average of the study area. Finally, as a validation method of the proposed methodology, the results of the frequency analysis were compared with resistivity data obtained with an electromagnetic conductivity meter, showing a very good correlation between the results.