Time-lapse crosshole GPR full-waveform inversion to characterize flow and transport processes in aquifers

Time-domain crosshole GPR full-waveform inversion has proven to be a powerful tool to characterize the subsurface and aquifers with a high resolution. The full-waveform inversion is able to provide both the relative permittivity and the electrical conductivity of the investigated medium and is therefore able to improve subsurface characterization. Until now, most of the studies have been performed in a steady state and correlations to flow and transport processes have been difficult. In this study, we investigate how the crosshole GPR FWI can be used to map and quantify different tracers in a gravel aquifer, that not only affect the permittivity of the subsurface, but also the electrical conductivity. Thereby, our main focus is to investigate the effect of a salt tracer using a realistic aquifer transport model. Synthetic time-lapse GPR data are used to analyze different full-waveform inversions and starting model strategies to find the method that reconstructs the tracer movement best. First experimental data using a positive salt tracer are investigated with a focus on the effective source wavelet estimation and optimization of the time-lapse full-waveform inversion results. First results show indicators for preferential flow paths and that most of the tracer travels faster at the bottom of the aquifer, while in the upper part a slow movement could be observed.