In situ 2-D and 3-D measurements of radiation patterns of half-wave dipole GPR antennas

Dipole antennas for ground-penetrating radar (GPR) radiate and receive electromagnetic waves with a strong directional dependence. Thus, experiments to measure in situ antenna radiation as functions of direction and polarization are of practical interest. Three field experiments were performed. One experiment was over a layered fluvial/eolian sequence; the other two used controlled targets (buried pipes and a metal ball). The radiation patterns were sampled by incrementally varying the antenna orientations and separations while recording reflections from the known targets. The results show qualitative, but systematic, correspondence with approximate theoretical far-field radiation patterns. Slow variations of amplitude with antenna azimuth and dip indicate that antenna orientations within 20° in the standard TE and TM acquisition geometries are adequate for most field applications, but not if detailed amplitude analysis is to be performed. Variations in antenna orientation or height (particularly for heights less than one-quarter wavelength) above the free surface introduce corresponding biases or uncertainties into recorded amplitudes. The variance within any suite of measurements is, in part, a consequence of differences in ground impedance at each antenna location. The theoretically predicted sensitivity to antenna height is mediated by surface roughness at high frequencies. It is necessary to include, or compensate for, the antenna radiation pattern in analysis of field data amplitudes, in experiment design, and in selecting appropriate antennas for specific applications.