Skip to Main Content
In this paper, we present the results of an on-going practical study into the development of novel FDTD numerical modelling and linear tomographic inversion methods for the interpretation and analysis of near-surface GPR data. The performance of these methods is evaluated against a simulated test-case example in which GPR data is collected over a realistic buried utility pipe model. The inversion approach is based on the Born approximation to the inverse-scattering problem and uses a Truncated Singular Value Decomposition (TSVD) to create the regularized solution. We asses whether informing the inversion by using a three-dimensional, full-field, O(2,4) accurate FDTD modelling scheme to calculate the incident field Green's functions improves the reconstruction outcome. By using a flexible modelling tool the antenna geometry, the air-ground interface and know subsurface artifacts are readily taken into account. The results show that the reconstruction images are made more accurate as the background scenario approached the (model) ground truth.
Date of Conference: 22-24 June 2011