[Thesis defence] 04/11/2024 - Quentin Didier: "Characterisation of natural environments on a macroscopic scale using mechanical and electromagnetic wave approaches". (UMR EMMAH)

Quentin Didier will publicly present his thesis entitled "Characterisation of natural environments on a macroscopic scale using mechanical and electromagnetic wave approaches. Application to the study of the impact of water content on the physical characteristics of a heterogeneous medium".

Laboratory

UMR 1114 EMMAH - Mediterranean Environment and Modelling of Agrohydrosystems

Composition of the jury

Ms GAELLE MESGOUEZ	Avignon University	Thesis supervisor
MR ROMAIN BROSSIER	Grenoble Alpes University	Rapporteur
Ms AMELIE LITMAN	Aix-Marseille University	Rapporteur
MR SLIMANE ARHAB	Avignon University	Thesis co-supervisor
MR JEAN-FRANCOIS SEMBLAT	ENSTA-ParisTech	Examiner
MR MOHAMMED SERHIR	Centrale-Supélec	Examiner
MR DOMINIQUE LESSELIER	Centrale-Supélec	Examiner

Summary

Knowledge of the structure and composition of the subsoil is a crucial issue for the coming decades. Indeed, climate change is leading to numerous droughts that could affect the water supply of populations. To better identify this resource, it is therefore important to develop non-destructive geophysical methods capable of accurately estimating the volume and distribution of groundwater. These methods must be able to distinguish water from other forms of subsoil heterogeneity. This thesis investigates different approaches for reconstructing the physical properties of the near-surface using geophysical data. Local optimisation algorithms have been identified as the most suitable for this purpose. The Gauss-Newton algorithm was chosen because it offers a very good compromise between performance and simplicity of implementation. Its robustness to noise and its spatial resolution were also tested in the context of an inversion of experimental microwave data. This doctoral work focuses in particular on the benefits of combining electromagnetic and mechanical geophysical data. These two types of measurements provide complementary information, as their sensitivities to soil properties are distinct. Mechanical data, for example, is mainly influenced by soil porosity, while electromagnetic data is more sensitive to water saturation. Inverting these data together can improve the accuracy of the reconstructions, taking advantage of the specific strengths of each method to fill in the gaps left by the other. In this study, particular attention was paid to integrating petrophysical relationships into the inversion algorithms. This constraint introduces a correlation between the various geophysical parameters to improve the quality of the reconstructions, but also to gain direct access to porosity and water saturation. In this context, various inversion strategies have been tested and compared in order to define the most appropriate for application to field data.

Keywords heterogeneous media, characterisation, inverse problems, mechanical waves, electromagnetic waves, rocky and agricultural soils