In-situ Characterization of the Infiltration Process using Time-Lapse Multi-Offset Gather Collected with an Array Antenna Ground Penetrating Radar

Abstract:

Recently, a family of array ground penetrating radars (GPR) has been developed. The basic concept is to electrically switch multiple antennas sequentially in several milliseconds to scan subsurface properties with minimal effort. The array GPR has been used to quickly generate 3D views of the subsurface in a wide range of applications, such as infrastructure inspections, archeological survey, or UXO and landmine detections. By placing the array GPR antenna at a fixed location, time-lapsed 3D GPR data can be easily collected with high reproducibility to visualize dynamic subsurface processes like vertical subsurface infiltration. In this study, a surface array ground penetrating radar (GPR) system of 3D-Radar AS was used to estimate the depth of the wetting front during an infiltration experiment conducted at an experimental field near the Tottori Sand Dune, Japan. The array GPR system used in this study consisted of horizontally aligned 10 transmitting (Tx) and 11 receiving (Rx) antennas. The system can scan 110 different antenna combinations in less than 1.5 seconds. The common-offset gather (COG) and common mid-point data (CMP) could be reconstructed from the scanned data. During the measurement, the array antenna position was fixed to ensure data reproducibility. In this study, the electromagnetic (EM) wave velocity structure was estimated from the reconstructed CMP data every 1 minute. Using the estimated EM wave velocity, the depth to the wetting front was computed. The estimated arrival time of the wetting front agreed well with the time when a sudden increase in the soil dielectric constant was independently measured with a soil moisture sensor at a depth below 20 cm. This study demonstrates that the array GPR system is capable of quantitatively and accurately tracking the infiltration front. Data from the array GPR system were then compared with the HYDRUS (2D/3D) simulation results to estimate field soil hydraulic parameters. This study was supported by JSPS Grant-in-aid Scientific Research Program (No. 25292147 and 16H02580) and by Joint Research Program of Arid Land Research Center, Tottori University.