Estimation of spatio-temporal variability of soil water content in agricultural fields with ground penetrating radar

Efficient water management, crop yield variability estimation and prediction of contaminant transport require some measurement of soil water content variation through time and space. This study focused on the estimation of spatio-temporal variability of volumetric soil water content (θv) in raised bed agricultural fields using ground penetrating radar (GPR), comparison of GPR method with gravimetric sampling data and development of 2D maps of θv. The GPR system (pulse EKKO Pro) with 200 MHz antennas was used to collect data on approximately 1.0 m wide and 13.0 m long raised beds of about 0.1 m height cultivated with vegetables. Transillumination Zero Offset Profile (Trans ZOP) and Transillumination Multiple Offset Gather (Trans MOG) GPR survey modes which are classically used as borehole GPR method were employed as a surface GPR method. In each of these survey modes, the direct ground wave travel time was measured. The θv at each Trans ZOP and Trans MOG location was calculated by first converting the electromagnetic (EM) wave velocity into soil dielectric permittivity and then to θv using a standard empirical relationship. The results revealed that the spatio-temporal variability of θv in raised bed agricultural fields could be estimated using the Trans ZOP and Trans MOG GPR survey modes. The GPR estimated θv and gravimetrically measured soil water content (θg) were not significantly different (P = 0.272). The correlation coefficient was 0.87, the root mean square error was 0.0184 m3/m3 and the average error was 0.20% between the two methods. The Trans MOG survey data allowed us to create plan view maps (2D) of the θv variation which could not be obtained from the Trans ZOP data. No statistical difference (P = 0.053) was found between the Trans ZOP and average Trans MOG values.