Assessing the Repeatability of GPR Data for Hydrological Monitoring Applications

Ground-penetrating radar (GPR) is a promising geophysical method for investigating subsurface flow processes with high spatial and temporal resolution. While several field examples of GPR monitoring using different measurement strategies have demonstrated its suitability for imaging subsurface processes, only little is known about the data quality and the data repeatability of such experiments. Therefore, we present a systematic field experiment, where we monitor nonuniform infiltration dynamics with repeated 3-D GPR measurements (4-D GPR). In a detailed analysis of the acquired 4-D GPR data, we focus on evaluating data repeatability as well as sources and magnitude of amplitude fluctuations and time shifts not related to subsurface flow processes (i.e., time-lapse noise). Our analyses show that we are able to quantify data repeatability and thus, to set thresholds for further data interpretation. Demonstrating the application of these thresholds, we perform a first-order, attribute-based interpretation of our 4-D GPR data. The results corroborate that we are able to collect temporally and spatially repeatable GPR data and allow us to provide recommendations for performing high-quality GPR monitoring experiments.