Calibrating the impact of root orientation on root quantification using ground-penetrating radar

Ground-penetrating radar (GPR) has provided a non-invasive means for field root investigation. However, the horizontal cross angle (x) of root orientation intersecting a survey line considerably impacts the amplitude area (A) reflected from a root and impairs the accuracy of GPR-based root quantification. Prediction of A(90A degrees) (the value of A scanning at x = 90A degrees) from multiple A(x) measurements could correct such impact. Previous method of A(90A degrees) prediction focused on target roots at field point scale. The aim of this study is to develop a method to predict A(90A degrees) at field plot scale. A(90A degrees) was predicted by a pair of A(x) measured at two arbitrary scanning lines together with an estimated soil background amplitude area. Three independent datasets were employed to test the proposed method. The field experiment included radar data collected for six roots of Caragana microphylla in a sandy-clay soil at four cross angles (30A degrees, 45A degrees, 60A degrees, and 90A degrees). The sand box experiment included radar data for 12 dowels at 13 cross angles (0A degrees to 180A degrees, in 15A degrees steps). The simulation experiment included A(x) of 46 simulated roots at 13 cross angles (0A degrees to 180A degrees, in 15A degrees steps). For all experiments, A(90A degrees) was accurately estimated. Root orientation could also be determined. After correcting the impact of cross angle, the accuracy of root diameter estimation improved. Correlation coefficient between actual and estimated root diameters increased from 0.77 to 0.81, with RMSE declining from 9.53 to 7.05 mm. A method of correcting the influence of root orientation on root GPR signal at the field plot scale has been established. This method enhances root quantification using GPR.