Non-invasive estimation of root zone soil moisture from coarse root reflections in ground-penetrating radar images

Abstract

Root zone soil moisture is an important component in water cycling through the soil-plant-atmosphere continuum. However, its measurement in the field remains a challenge, especially non-invasively and repeatedly. Here, we developed a new method that uses ground-penetrating radar (GPR) to quantify root zone soil moisture. Coarse roots were chosen as reflectors to collect GPR radargrams. An automatic hyperbola detection algorithm identified coarse root reflections in GPR radargrams and determined the velocity of GPR wave, which then was used to calculate the average soil water content of a soil profile (ASWC) and soil water content in a depth interval (ISWC). In total, GPR reflection data of 55 root samples from three computer simulation scenarios and two field experiments in sandy shrubland, one burying roots at known depths and the other under the undisturbed condition, were used to evaluate the proposed method. Both the simulated and the field collected data demonstrated the effectiveness of the proposed method for measuring root zone soil moisture with high accuracy. Even in the two field experiments, the root-mean-square errors of the estimated ASWC and ISWC relative to measurements from soil cores were as low as 0.003 and 0.012 m3·m−3, respectively. The proposed method offers a new way of quantifying root zone soil moisture non-invasively that allows repeated measurements. This study expands the application of GPR in root and soil study and enhances our ability to monitor plant-soil-water interactions.