Realistic model of dispersive soils using PLRC-FDTD with applications to gpr systems

Abstract

A realistic model of ground soil is developed for the electromagnetic simulation of Ground Penetrating Radar (GPR) systems. A three dimensional Finite Difference Time Domain (FDTD) algorithm is formulated to model dispersive media using N-term Debye permittivity function with static conductivity. The formulation of the algorithm is based on the concept of the Piecewise Linear Recursive Convolution (PLRC) in order to simulate the dispersion properties of soil as a two-term Debye medium. This approach of ground modeling enhances the accuracy and reliability of results obtained for GPR problems. The developed algorithm is validated when simulating practical GPR Systems used to detect different objects buried in Puerto-Rico and San Antonio clay loams. The proposed algorithm is employed to compare the impact of using two-term Debye model to simulate real soil on the coupling coefficient between transmitting and receiving antennas due to the absence and presence of buried targets to that of using non-dispersive soil model. The effect of soil moisture content on the performance of GPR system in detecting buried objects such as metallic and plastic pipes is investigated.