Computing localized breakthrough curves and velocities of saline tracer from ground penetrating radar monitoring experiments in fractured rock

Solute tracer tests are an established method for the characterization of flow and transport processes in fractured rock. Such tests are often monitored with borehole sensors which offer high temporal sampling and signal to noise ratio, but only limited spatial deployment possibilities. Ground penetrating radar (GPR) is sensitive to electromagnetic properties, and can thus be used to monitor the transport behavior of electrically conductive tracers. Since GPR waves can sample large volumes that are practically inaccessible by traditional borehole sensors, they are expected to increase the spatial resolution of tracer experiments. In this manuscript, we describe two approaches to infer quantitative hydrological data from time-lapse borehole reflection GPR experiments with saline tracers in fractured rock. An important prerequisite of our method includes the generation of GPR data difference images. We show how the calculation of difference radar breakthrough curves (DRBTC) allows to retrieve relative electrical conductivity breakthrough curves for theoretically arbitrary locations in the subsurface. For sufficiently small fracture apertures we found the relation between the DRBTC values and the electrical conductivity in the fracture to be quasi-linear. Additionally, we describe a flow path reconstruction procedure that allows computing approximate flow path distances using reflection GPR data from at least two boreholes. From the temporal information during the time-lapse GPR surveys, we are finally able to calculate flow-path averaged tracer velocities. Our new methods were applied to a field data set that was acquired at the Grimsel Test Site in Switzerland. DRBTCs were successfully calculated for previously inaccessible locations in the experimental rock volume and the flow path averaged velocity field was found to be in good accordance with previous studies at the Grimsel Test Site. ; ISSN:1996-1073