Microwave tomography for subsurface prospecting

Ground Penetrating Radar (GPR) is a diagnostic tool that well assessed in a variety of areas, including geophysics, archaeological prospection, civil engineering, and planetary exploration, just to mention few examples.

Notwithstanding the simplicity of the underlying principle, a significant limitation of GPR is concerned with the interpretability of the raw data, mostly presented under the form of radargrams, particularly in scenarios that are complicated and characterised by a multitude of embedded targets. 

To enhance the interoperability of radar images, it is crucial to reliably model the electromagnetic scattering, so that the radar imaging is conceptualised as an inverse scattering problem. For such an inverse problem, the geometric and electromagnetic properties of the targets are retrieved by the field scattered by the target when an incident field impinges on it. Despite the simplicity of the underlying principle, this inverse problem is inherently complex due to its non-linear nature and ill-posedness. These mathematical difficulties have a detrimental effect on the effectiveness of GPR diagnostics in real cases. To facilitate the application of inverse scattering approaches in real-world scenarios, it is necessary to resort to approximate models of electromagnetic scattering. Microwave tomography exploits the linearization of the inverse scattering problem, so that reconstruction approaches can be developed that operate under realistic conditions with the aim of estimating the position and geometry of targets, albeit with limitations on the class of unknowns that can be reconstructed (i.e. resolution limitations) and the impossibility of quantitatively estimating the electromagnetic properties of targets.

In the talk, microwave tomography will presented under a unified mathematical framework based on the solution of an integral equation accounting for arbitrary measurement configurations and background scenarios for both contact and contactless GPR measurements. Furthermore, the investigation of the reconstruction performance of the microwave tomographic approach for different measurement configurations and background scenarios will be presented.

Finally, several cases of exploitation of the microwave tomographic approach in real cases will be shown.