A new drone-borne GPR for soil moisture mapping

Wu, Kaijun [UCL] Rodriguez, Gabriela Arambulo [UCL] Zajc, Marjana [UCL] Jacquemin, Elodie [UCL] Clément, Michiels [UCL] De Coster, Albéric [UCL] Lambot, Sébastien [UCL]

In this study, we set up a new drone-borne ground-penetrating radar (GPR) for soil moisture mapping. The whole radar system weighs 1.5 kg and consists of a handheld vector network analyzer (VNA) working as frequency domain radar, a lightweight hybrid horn-dipole antenna covering a wide frequency range (250–2800 MHz), a GPS for positioning, a microcomputer with the controlling application, and a smartphone for remote control. Soil moisture is derived from the radar data using full-wave inverse modeling based on the radar equation of Lambot et al. and multilayered media Green's functions. The inversion is performed in the time domain and focuses on the surface reflection. The antenna-drone system is characterized by global reflection and transmission functions which are determined through a calibration procedure. We performed drone-GPR measurements over three different agricultural fields in the loess belt region of Belgium. In this study, we used the 500–700 MHz range to avoid soil surface roughness effects and to focus on the top 10–20 cm of the soil. These fields present a range of landform conditions leading to specific soil moisture distributions. The soil moisture maps were constructed from the local measurements using kriging. The obtained soil moisture maps are in good agreement with the topographical conditions of the fields and aerial orthophotography observations. These results demonstrated the potential and benefits of drone-GPR for fast, high-resolution mapping of soil moisture at the field scale, and to support, e.g., precision agriculture and environmental monitoring.