| Ground penetrating radar uses antennas to transmit electromagnetic energy into the subsurface and record energy scattered from subsurface objects. Antenna patterns describe the amplitude and vibration directions of the electromagnetic fields radiated by antennas. Radiation patterns are a function of many variables such as frequency, observation distance, soil electrical properties, antenna design, and antenna height above the interface. Knowledge of how these properties affect GPR antenna patterns are important for the proper design of ground penetrating radar surveys and data interpretation. Antenna pattern information can be combined with the scattering characteristics of subsurface objects to constrain the size, shape, orientation, and physical properties of buried objects.; Numerical modeling of antenna patterns provides physical insight into radiation mechanisms and the effects of physical properties on antenna patterns. Wave type, polarization, and amplitudes of radiated fields are investigated for a variety of soil properties and antenna heights relative to the air-soil interface. Antenna patterns are not commonly utilized in GPR because near-field patterns are poorly understood and because they are a function of frequency, soil properties, and depth of investigation. Traditional asymptotic, geometrical optics antenna pattern solutions neglect the lateral wave term and are insufficient for many ground penetrating radar (GPR) applications. Space and lateral wave mechanisms are clearly observed in numerical simulations using finite difference time domain (FDTD) models. Numerical models of antenna patterns are verified using physical experiments over a water filled tank.; The polarization dependent scattering characteristics of planes and cylinders, that represent such commonly encountered objects as stratigraphy and buried utilities, are used to illustrate the significance of polarization for imaging subsurface objects. Analytical solutions are plotted for a variety of different impedance contrasts and incidence angles, and the results are analyzed and summarized to illustrate important polarization dependent scattering characteristics and trends. While these analytical solutions require minimal computational resources and provide valuable insight, they contain simplifying assumptions. The important polarization dependent scattering trends observed from the analytical solutions are verified with physical experiments and FDTD simulations that are very intensive in both computer random access memory (RAM) and central processing unit (CPU) time, but do not contain these far-field assumptions and work well for near-surface simulations. Finally, the affects of antenna patterns on data processing are investigated after studying antenna radiation properties and how they affect scattering from subsurface objects. |
