Non-invasive electromagnetic methods are increasingly utilized in geophysical engineering, infrastructure characterization, and environmental studies. Various geophysical techniques estimate medium properties, monitor soil conditions, and provide crucial soil water content and hydraulic parameters for understanding dynamic hydrological processes. Traditionally, soil water content estimates rely on subsurface permittivity and conductivity, using models like the Complex Refractive Index Model (CRIM) or empirical equations such as Topp's and Archie's law. Surface ground penetrating radar (GPR) is particularly effective for mapping subsurface dielectric permittivity. However, conventional ray-based techniques for conductivity estimation often yield significant errors due to their reliance on limited data and simplified models. Full-waveform inversion (FWI) addresses these challenges by employing accurate forward modeling, utilizing a comprehensive dataset to provide reliable estimates of permittivity and conductivity. This work presents a novel FWI scheme that estimates these values from surface GPR data, based on a frequency-domain solution of Maxwell’s equations in a three-dimensional, horizontally layered subsurface model. The method updates permittivity and conductivity alongside the source wavelet's phase and amplitude, optimizing medium properties through a gradient-free approach. Applied to ground and reflected wave analysi
