Douglas S. Sassen

3D polarimetric GPR coherency attributes and full-waveform inversion of transmission data for characterizing fractured rock

Ground-penetrating radar (GPR) can detect and describe fractures to help us characterize fractured rock formations. A fracture alters the incident waveform, or wave shape, of a GPR signal through constructive and destructive interference, depending on the aperture, fill, and orientation of the fracture. Because the electromagnetic (EM) waves of GPR are vectorial, features exhibiting strong directionality can change the state of polarization of the incident field. GPR methods that focus on changes in waveform or polarization can improve detection and discrimination of fractures within rock bodies. An algorithm based on coherency, a seismic attribute that delineates discontinuities in wavelet shape, is developed for polarimetric GPR. It uses the largest eigenvalue of the time-domain scattering matrix when calculating coherence. This algorithm is sensitive to wave shape and is unbiased by the polarization of GPR antennas. Polarimetric coherency works better than scalar coherency in removing the effects of polarization on field data collected from a fractured limestone plot used for hydrologic experimentation. Another method, for time-domain full-waveform inversion of transmission data, quantitatively determines fracture aperture and EM properties of fill, based on a thin-layer model. Inversion results from field data show consistency with the location of fractures from reflection data. These two methods offer better fracture-detection capability and quantitative information on fracture aperture, dielectric permittivity, and electrical conductivity of the fill than traditional GPR imaging and scalar-coherency attributes.

Ecohydrogeophysics at the Edwards Aquifer: insights from polarimetric ground-penetrating radar

Three-dimensional multicomponent ground-penetrating radar (GPR) reflection data and horizontal GPR transmission profiles were acquired and analyzed to better understand the interaction of vegetation with subsurface flow conduits at a hydrologic experimentation site. Previous researchers conducted a set of shallow ( Juniperus ashei brush control on the local hydrology. Tracer experiments showed a high degree of variability in tracer recovery, advection speed and concentration depending on the location of the application of the tracer. Both 3D multicomponent GPR reflection images and coherency and inversion of GPR horizontal transmission profiles were utilized to identify the main conduits of flow within the experimentation site in order to explain the observations of the experiments. The 3D multicomponent GPR and coherency images revealed that the most obvious potential conduits run nearly parallel with the observation trench. Inversions of the horizontal transmission profiles indicate that some conduits are filled with soil while others have no fill. This information helps explain the high spatiotemporal variability observed in the tracer data. Additionally, the GPR and hydrologic experiments suggest that Juniperus ashei significantly impacts infiltration by redirecting flow towards its roots occupying fractures within the rock. This study demonstrates that GPR provides a noninvasive tool that can improve future subsurface ecohydrologic experimentation.

Coherency Attribute Algorithm For Polarimetric Ground Penetrating Radar (GPR)

Coherency is a traditionally seismic attribute used for the delineation of discontinuities in wavelet shape. Coherency attributes compare a small window of a wave trace with the surrounding traces to determine the degree of similarity. While difference in the waveform data may be difficult to see in a migrated image, the coherency image will make it relatively easy to visualize. Coherency has been shown to be very useful in delineating faults and paleochannels in seismic data. In GPR changes in electric impedance, and target geometry affect the electromagnetic (EM) wavelet in a similar manner to seismic waves. Additionally, received EM backscatter from subsurface features depends strongly on the polarization of the transmitting and receiving antennas and the geometry and electrical properties of the scatterer. The EM waves of GPR are vector in nature with the electric and magnetic fields transverse to the direction of propagation. Features exhibiting strong directionality such as faults, fractures and edges can change the state of polarization of an incident field. Thus, methods that combine both polarization invariant information and wavelet shape information into a single, easily interpretable attribute image are desirable for the detection of discontinuities. To this end, the largest eigenvalue of the scattering matrix, acquired from polarimetric GPR, is used in conjunction with the eigenstructure coherency algorithm of Gersztenkorn and Marfurt (1996) to produce a coherency image unbiased by antenna polarization. The polarimetric coherency algorithm was applied to polarimetric GPR acquired over fractured and karsted limestone. The field results show significant improvement over scalar based coherency images.

3-D Multicomponent GPR to Characterize Shallow Subsurface Flow Paths In an Epikarst Limestone, Edwards Plateau, Texas

We acquired and analyzed 3-D multicomponent groundpenetrating radar (GPR) data to better understand the subsurface flow conduits at a hydrologic experimentation site. Previous workers conducted a set of shallow (< 2.5 m) subsurface hydrology experiments during simulated rainfall events within fractured and karsted limestone of the Edwards Aquifer region near San Antonio, Texas. They observed at an observation trench located on the down slope side of the site that lateral subsurface flow is guided by open joints, bedding planes and karst features. They also utilized tracer experiments, which showed a high degree of variability in tracer recovery, advection speed, and concentration depending on the location of the application of the tracer. We utilized the 3-D multicomponent GPR data in an attempt to identify the main conduits of flow within the experimentation site in order to explain the observed tracer experiments. The GPR revealed that the most obvious conduits run nearly parallel with the observation trench, with some conduits able to quickly move water away from the trench. This information helps explain the high spatiotemporal variability in the tracer data. Our study demonstrates that multicomponent GPR provides a technique that can improve future subsurface hydrologic experimentation.