Full-wave inversion of ground-penetrating radar data for forest litter characterization

Forest soil organic horizons constitute a major component of forest ecosystems and their accurate characterization is of prime importance for ecological studies as well as for carbon cycle and global change related studies. In other respects, the presence of forest litter is known to influence remote sensing radar data over forested areas and precise determination of litter radiative properties is necessary for proper processing of these data. In the present study, ultra wideband (0.8-4.0 GHz) ground-penetrating radar (GPR) data were collected above the forest floor of a beech forest with different litter layer thicknesses so as to examine the effect of litter on the backscattered radar signal and to investigate the potentialities of GPR for reconstructing litter constitutive properties. Full-wave inversion was used to process the radar data. Attenuation of the radar signal was found to increase as both operating frequency and litter thickness increase, as a result of the occurrence of dielectric and scattering losses within litter. Frequency dependence of the apparent electrical conductivity of litter was considered in the radar model to account for these phenomena. Close correspondence was observed between estimated and measured litter thicknesses and signal inversions provided reliable estimates of litter electromagnetic properties. These results show promising potentialities of the GPR technique for providing accurate and non-invasive characterization of forest litter.