On the application of the Stockwell transform to GPR data analysis

For most archaeological prospections with GPR, target detection is the most important aspect of the surveys. Shallow and narrow features may require a very fine profiles density to get detected. The smallest detectable size of archaeological materials is also dependent on the wavelength of the signal transmitted by the antenna. A buried archaeological debris field, containing higher concentrations of small fragments of ancient habitation, e.g. flint chips or other materials, may show some scattered energies on the radargrams which may slightly differ with respect to areas where no debris are located. Unless a fortunate color transform is used, it is unlikely that these small reflections will ever be noticed within the radargram - vertical slice - dataset [1], [2]. Rather than showing the changes in reflected energies along the vertical - radargram - slices of the ground, it is often more useful to map horizontal changes of reflected energy across a site. In this case small but consistent reflections above the background noise can be visualized by using time slice analysis. Some of the main goals of GPR data processing are: to filter the radar section from clutters and out of band noises; to enhance coherent reflections likely due to the target response; to transform the acquired time section into a depth section; to build a 3D image of the reflection targets in the subsoil by correlating reflections from adjacent radargrams [1], [2]. In this paper we show some application of the Stockwell transform, both to synthetic and experimental data, to enhance the signal to noise ratio in radargrams.