Validation of ground penetrating radar full-waveform inversion for field scale soil moisture mapping

Abstract

Ground penetrating radar (GPR) is an efficient method for soil moisture mapping at the field scale, bridging the scale gap between small-scale invasive sensors and large-scale remote sensing instruments. Nevertheless, commonly-used GPR approaches for soil moisture characterization suffer from several limitations and the determination of the uncertainties in GPR soil moisture sensing has been poorly addressed. Herein, we used a proximal GPR method based on full-waveform inversion of ultra-wideband radar data for mapping soil moisture and we evaluated uncertainties in the soil moisture maps by three methods. First, GPR-derived soil moisture uncertainties were computed from GPR data inversions, according to measurements and modeling errors, and to the sensitivity of the electromagnetic model to soil moisture. Second, the repeatability of soil moisture mapping was evaluated. Third, GPR-derived soil moisture was compared with ground-truth measurements (soil core sampling). The proposed GPR method appeared to be highly precise and accurate, with a spatially averaged GPR inversion uncertainty of 0.0039 m3 m-3, a repetition uncertainty of 0.0169 m3 m-3, and an uncertainty of 0.0233 m3 m-3 when compared with ground-truth measurements. These uncertainties were mapped and appeared to be related to some local model inadequacies and to small-scale variability of soil moisture. In a soil moisture mapping framework, the interpolation was found to be the main source of the observed uncertainties. The proposed GPR method was proven to be largely reliable in terms of accuracy and precision and appeared to be highly efficient for soil moisture mapping at the field scale.