Surface Roughness Clutter Effects in GPR Modeling and Detection

Abstract

Computational modeling of ground penetrating radar (GPR) wave propagation in aidsoil and scattering from buried dielectric targets shows that the largest source clutter is due to the rough ground interface. Conclusions are drawn from the observed features that are somewhat unexpected. One important observation is that while target resolution increases with increasing frequency, the target features are harder to separate from the background clutter of the rough ground interface. The limitation on frequency is not the lack of penetration depth in lossy soil, but rather the greater phase effects of ground surface dips and bumps. Similarly, a greater soil moisture level increases the wave decay rate, but it is the reduction in wavelength that makes the rough surface appear electrically larger, which increases clutter and makes the target harder to resolve. Means of minimizing the rough surface clutter distortion include active ground surface subtraction in the time domain and convolution with a delayed time window in the frequency domain.