Quantifying field-scale surface soil water content from proximal GPR signal inversion in the time domain
K.Z. Jadoon, S. Lambot, B. Scharnagl, J. van der Kruk, E. Slob and H. Vereecken
Issue: Vol 8, No 6, December 2010 pp. 483 - 491
Info: Article, PDF ( 3.51Mb )
We applied inverse modelling of zero-offset, air-raised ground-penetrating radar (GPR) data to measure soil surface water contents over a bare agricultural field. The GPR system consisted of a vector network analyser combined with a low-frequency 0.2–2.0 GHz off-ground monostatic horn antenna, thereby setting up an ultra-wideband stepped-frequency continuous-wave radar. A fully automated platform was created by mounting the radar system on a truck for real-time data acquisition. An antenna calibration experiment was performed by lifting the whole setup to different heights above a perfect electrical conductor. This calibration procedure allowed the flittering out of the antenna effects and antenna-soil interactions from the raw radar data in the frequency domain. To avoid surface roughness effects, only the lower frequency range of 0.2–0.8 GHz was used for signal processing. Inversions of the radar data using the Green’s functions were performed in the time domain, focusing on a time window containing the surface reflection. GPR measurements were conducted every 4 m along a transect of 100 m. In addition, five time-domain reflectometry measurements were randomly recorded within the footprint of the GPR antenna. A good agreement was observed between the GPR and time-domain reflectometry soil water content estimates, as compared to the previous study performed at the same test site using a higher frequency 0.8–1.6 GHz horn antenna. To monitor the dynamics of soil water content, a pair of time-domain reflectometry probes was installed at 8 cm depth near the footprint of the GPR antenna and both time-domain reflectometry and GPR measurements were carried out for a period of 20 days. A good agreement of the trend was observed between the time-domain reflectometry and GPR time-lapse data with respect to several precipitation events. The proposed method and truck-mounted setup appear to be promising for the real-time mapping and monitoring of surface soil moisture contents at the field scale.