Fresnel reflection coefficients for GPR‐AVA analysis and detection of seawater and NAPL contaminants

Abstract

We obtain the transverse electric (TE) and transverse magnetic (TM) Fresnel reflection coefficients for different interfaces in the subsoil: air/fresh‐water, air/seawater, fresh‐water/seawater, air/NAPL (non‐aqueous phase liquid), NAPL/water and water/NAPL. We consider a range of NAPL saturations, where the complementary fluid is water with 0.65 ppt (parts per thousand) of NaCl. The common feature is that the TM mode (parallel polarization) has a negative anomaly and the TE mode (perpendicular polarization) has a positive anomaly. For the cases studied in this work, pseudo‐Brewster angles appear beyond 40° for the air/NAPL and NAPL/water interfaces and at near offsets (below 40°) for the water/NAPL interface. Pseudo‐critical angles are present for the water/NAPL interface. Besides the reflection strength, the phase angle can be used to discriminate between low‐ and high‐conductivity NAPL, when the properties of the upper medium are known. A wavenumber–frequency domain method is used to compute the reflection coefficient and phase angle from synthetic radargrams. This method and the curves can be used to interpret the amplitude variations with angle (AVA) of reflection events in radargrams obtained with ground‐penetrating radar (GPR).

On the use of dispersive APVO GPR curves for thin-bed properties estimation: Theory and application to fracture characterization

“…EM reflection properties, which have been widely studied for GPR purposes ͑Annan, 2001; Grégoire et al, 2003;Carcione et al, 2006͒, are sensitive to several contributing properties such as dielectric permittivity, electrical conductivity, and magnetic permeability ͑Annan, 2001͒. They also depend on the antenna polarization ͑Leh-mann, 1996; Lutz et al, 2003͒, on the incident angle of the GPR wave ͑Annan, 2001͒, and on the frequency of the studied wave.…”

“…These variations were related to lithological ͑Kindelan et al, 1989͒ and fluid content changes ͑Simmons and Backus, 1994;Hall and Kendall, 2003;Mahob and Castagna, 2003;Stovas et al, 2006͒. AVO tests were successfully performed on GPR data to qualitatively characterize the presence of nonaqueous-phase liquid ͑NAPL͒ contaminants in the subsurface ͑Baker, 1998; Deeds and Bradford, 2002;Jordan and Baker, 2002;Jordan et al, 2004͒. Work was supplemented by numerical analyses of Lehmann ͑1996͒ andCarcione et al ͑2006͒, who studied transverse electric ͑TE͒ and transverse magnetic ͑TM͒ reflection variations with offset for different contrasts of EM properties such as NAPL concentrations. All these studies were based on the Fresnel reflection coefficient curves ͑Griffiths, 1998͒, which were derived assuming monochromatic EM incident plane waves on a boundary separating two homogeneous and isotropic media.…”

GPR and high resolution seismic integrated methods to understand the liquefaction phenomena in the Mirabello Village (earthquake ML 5.9, 2012)

“…Another important factor in the investigation of subsurface structures is the shape of the lobes of electromagnetic energy propagation, and therefore how electromagnetic waves change in the near field (e.g., Yarovoy et al, 2007). To better understand the location, size, and orientation of sand-filled feeder dikes, we collected GPR data by towing the antenna using two orientations: transverse and parallel modes (Carcione et al, 2006). This was done because the quality of the imaging of the subsoil depends on how the electromagnetic radiation incides on the liquefaction dikes.…”

Fracture aperture and fill characterization in a limestone quarry using GPR thin‐layer AVA analysis

“…bright or dim spots in CO sections) can show very different behaviour when examined by AVA analysis. Carcione et al, (2006) studied synthetic AVA curves for various contrasts in EM properties; they highlighted the potential of AVA analysis for characterizing non aqueous phase liquid (NAPL) contamination considering a single interface between an uncontaminated upper layer and a lower layer contaminated with LNAPL. Applying the analysis to GPR data, Deeds and Bradford (2002) successfully characterized the presence of an NAPL contaminated zone at an alluvium/clay boundary.…”