December 01, 2022
16h
Sud 03
Characterization of soil electromagnetic properties has always been an important task for hydrology, meteorology, environmental sciences, agriculture etc. In the meanwhile, airborne technology has been developed and applied to various applications.
This thesis aimed to develop and analyse new ground-penetrating radar (GPR) techniques that are specifically designed for Unmanned Aerial Vehicles (UAV). The thesis involves radar antenna design and calibration, frequency sensitivity analyses based on the reflexion coefficient and full-wave inversion, numerical simulations using analytical Green’s functions and finite-difference time-domain (FDTD) methods, and validations with field measurements. The concept that drone-borne GPR and the full-wave inversion methods can be used for soil moisture and electrical conductivity mapping using drone was demonstrated. In particular, at relatively low frequencies, i.e., < 50 MHz, the soil surface reflexion coefficient becomes strongly sensitive to electrical conductivity and much less sensitive to permittivity. We extended the radar equation and its calibration to varying antenna pointing angles in both the E- and H-planes, which is of concern for UAV-GPR applications. Finally, we analysed the reflection problem for the trench-hill structures of the potato fields in order to map the root-zone soil moisture, e.g., to support irrigation practices.
The outcomes of this thesis are useful in precision agriculture for sustainable production, by permitting real-time determination for precise and automatic irrigation. They enrich air-coupled GPR application avenues for soil surveys in terms of the root-zone electrical conductivity, by operating at intermediate frequencies between GPR and electromagnetic induction (EMI). They enrich remote sensing methods and applications of drone technology in the soil mapping aspect, by providing more information on the soil properties and giving soil maps with higher resolution and efficiency.