High-Resolution Time-Lapse Monitoring of Unsaturated Flow using Automated GPR Data Collection

Abstract:

High-resolution ground-penetrating radar (GPR) data provide the detailed information required to image subsurface structures. Recent advances in GPR monitoring now also make it possible to study transient hydrologic processes, but high-speed data acquisition is critical for this application. We therefore highlight the capabilities of our automated system to acquire time-lapse, high-resolution multifold GPR data during infiltration of water into soils. The system design allows for fast acquisition of constant-offset (COP) and common-midpoint profiles (CMP) to monitor unsaturated flow at multiple locations. Qualitative interpretation of the unprocessed COPs can provide substantial information regarding the hydrologic response of the system, such as the complexities of patterns associated with the wetting of the soil and geophysical evidence of non-uniform propagation of a wetting front. While we find that unprocessed images are informative, we show that the spatial variability of velocity introduced by infiltration events can complicate the images and that migration of the data is an effective tool to improve interpretability of the time-lapse images. The ability of the system to collect high density CMP data also introduces the potential for improving the velocity model along with the image via reflection tomography in the post-migrated domain. We show that for both simulated and empirical time-lapse GPR profiles we can resolve a propagating wetting front in the soil that is in good agreement with the response of in-situ soil moisture measurements. The data from these experiments illustrate the importance of high-speed, high-resolution GPR data acquisition for obtaining insight about the dynamics of hydrologic events. Continuing research is aimed at improving the quantitative analysis of surface-based GPR monitoring data for identifying preferential flow in soils.