Subsurface permittivity estimation from ground-penetrating radar measurements

Ground penetrating radar (GPR) has been the focus of a surge of research activity during the last ten years, largely due to its utility in applications such as geological and civil surveying, and locating buried ordnance. The recent development of low cost, low power pulse-echo scanned GPR systems has also made GPR systems affordable imaging tools. One conventional method for acquiring subsurface GPR data is to scan an antenna array over the region to be probed to acquire monostatic or multistatic space-time echoes of the subsurface scatterers. The data can then be processed to form an easily interpretable twoor three-dimensional image of the scanned region. Many algorithms for imaging objects in a weakly inhomogeneous dielectric halfspace using microwaves have been studied. These methods are applicable to GPR since many subsurface scenes may be approximately modeled as perturbations interspersed throughout clutter in an otherwise homogeneous background. The effectiveness of each imaging method depends on the accuracy of the estimate of the electromagnetic wave velocity in the background material. We propose a method for estimating the ground´s background velocity which accounts for an air-gap between the aperture and the air-ground interface, as in reflection coefficient measurement, but which does not require a plate calibration. This method is an autofocusing technique which exploits Fourier processed backscatter from point scatterers buried beneath the surface. Unlike time-of-flight methods, it also accounts for diffractive effects