Numerical simulations and analysis for airborne ground penetrating radar

We conduct numerical simulations for an airborne ground penetrating radar (GPR) to investigate its performance in different situations such as antenna configuration and ground surface roughness with 3-dimensional (3D) finite-difference timedomain (FDTD) method. The bow-tie antennas are used in the simulations. Gaussian random rough surface is used to simulate ground surface and underground interfaces. We have investigated four antenna configurations (FBx, FBy, LRx, and LRy) in line with the configuration of airborne GPR survey, and analyzed correspondent simulated profiling results. It is concluded that LRx is the most optimal way for 3D experiments. Antenna height is an important factor in airborne GPR survey. As the antenna height increases, the reflected energies from both underground interfaces and the ground surface decrease, but the ratio between them increases. The roughness of the ground surface affects the duration of the coda waves resulted from surface reflection. The detection of underground features heavily depends on the separation of the signature of rough surface and the targets in radar profile. It is concluded that airborne GPR is a potential and optimistic tool from the numerical experiments.