2D and 3D ground‐penetrating radar surveys with a modular system: data processing strategies and results from archaeological field tests. Issue 2 (1st January 2013)

• Record Type: Journal Article
• Authors: Verdonck, Lieven; Vermeulen, Frank; Docter, Roald; Meyer, Cornelius; Kniess, Rudolf
• Is Part Of: Near surface geophysics. Volume 11:Issue 2(2013)
• Page Start: 239
• Page End: 252
• Publication Date: 2013-01-01
• Subjects: Earth (Planet) -- Surface -- Periodicals; Geophysics -- Technique -- Periodicals; Engineering geology -- Periodicals; Geophysics -- Periodicals; Planets -- Surfaces; Engineering geology; Geophysics -- Technique; Geophysics; Earth (Planet); Periodicals; 550
• DOI: 10.3997/1873-0604.2013007
• Languages: English
• ISSNs: 1569-4445

Abstract

ABSTRACT: Recently, the use of ground‐penetrating radar (GPR) arrays with a large number of antenna elements in a fixed configuration has become more common. The investment needed for these systems is significant. In order to reduce the recording time in the field, an alternative is the use of several single GPR antennas in parallel (a 'modular system'). Although this does not match the fast acquisition of detailed data sets by means of multi‐channel arrays, a system consisting in single antennas can gradually be expanded and investment can be spread over time. This paper presents a 2D and a full‐resolution 3D survey, conducted with a modular GPR instrument. A characteristic of these systems is that the cross‐line separation between transmitter‐receiver pairs is larger than the sampling distance prescribed by the Nyquist theorem. As a consequence, for 3D data collection, profiles have to be acquired between previously recorded ones, which requires high positioning accuracy. A completely identical response for different single GPR antennas is difficult to achieve. For the system tested, on less favourable soils this resulted in striping in the horizontal slices. Several methods (3D frequency‐wavenumber filtering, eigenimage filtering, mean profile filtering and filtering based on discrete wavelet transform, discrete ridgelet transform and linear Radon transform) were applied to two data sets exhibiting different kinds of linear noise and their capability to suppress artefacts was assessed. Although overall a reduction of the stripe patterns was achieved, mostly it was impossible to fully eliminate the noise in the time‐slices without low‐pass filtering in the cross‐line direction. For the 2D data, low‐pass filtering caused loss of some of the archaeological response and therefore was not applied. Mean profile filtering allowed the most reliable characterization of the archaeological structures.