Sub-surface or ground penetrating radar (GPR) refers to the wide range of techniques designed for locating objects or interfaces beneath the surface of the earth. These techniques are also relevant to the more general application of remotely investigating any solid or liquid dielectric region and the space obscured by such a region.
The overall design philosophy of the methodology applied, and also the specific detail of the technique depend on the type of target sought and the nature of the obscuring medium. The types of system employed can be either local to the surface to be probed, or can be remote, for example airborne or satellite systems. The successful deployment of a system depends on a number of factors that must be incorporated into the design including electromagnetic considerations, soil science, geophysics and signal processing.
Because of the complexity of the design process for the electromagnetic system, the design of GPR has progressed empirically through experimentation. The usual domain decomposition methods applied in EM design activities cannot be applied in the case of GPR because of the proximity of antenna, ground and target. This means that the response of any component in the system dependents on all other components. History shows that empirical approaches to problem solving provide solutions, but they may not be optimal. Furthermore, this work is complex, time consuming and expensive.
An alternative that has recently become available is numerical modelling. In this approach the initial design and optimisation of the electromagnetic system are carried out using computer simulation. Accurate models of the antenna, ground and scatterer are constructed and the system response evaluated. This alternative is fast, relatively cheap and allows a greater breadth of consideration in the scoping and optimisation of systems. This approach is relatively recent due to the difficulty of representing the complex electromagnetic environment using computer models.
The work carried out within our company is at the forefront of numerical modelling techniques for the computational investigation of GPR system performance. Based on the finite difference time domain method (FDTD), the software and expertise that we have can provide solutions for a wide range of soil media, frequencies and bandwidths, and antenna types from simple CW systems to pulsed wide band horns and loaded dipoles.
In the following pages a brief introduction is provided on GPR systems, their primary components and related design issues.
The impact of computational modelling in design and the benefits that modelling provides over conventional empirical design philosophies is discussed here.