Reflection waveform inversion of ground-penetrating radar data for characterizing thin and ultrathin layers of nonaqueous phase liquid contaminants in stratified media

Abstract

Accurately quantifying thin-layer parameters by applying a targeted reflection waveform inversion methodology to ground-penetrating radar (GPR) reflection data may provide a useful tool for near-surface investigation and especially for contaminated site investigation where nonaqueous phase liquid (NAPL) contaminants are present. We implemented a targeted reflection waveform inversion algorithm to quantify thin-layer permittivity, thickness, and conductivity for NAPL thin (≤ 1∕2 dominant wavelength λ) and ultrathin (≤ 1∕8λ) layers using GPR reflection data. The inversion used a nonlinear grid search with a Monte Carlo scheme to initialize starting values to find the global minimum. By taking a targeted approach using a time window around the peak amplitude of the reflection event of interest, our algorithm reduced the complexity in the inverse problem. We tested the inversion on three different synthetic data sets and four field data sets. In all testing, the inversion solved for NAPL-layer properties within 15% of the measured values. This algorithm provides a tool for site managers to prioritize remediation efforts based on quantitative assessments of contaminant quantity and location using GPR.

A synthetic study to assess the applicability of full-waveform inversion to infer snow stratigraphy from upward-looking ground-penetrating radar data

“…Because upGPR considers only vertical incident wavepaths, the antenna radiation patterns do not influence the waveforms. Babcock and Bradford (2015) show that the plane-wave approximation produced similar data as obtained with spreading-corrected simulations using a 2D finite-difference time-domain forward algorithm. The plane-wave approach violates the far-field condition within one to three wavelengths.…”

“…For the sake of simplicity, we make a plane-wave assumption. As discussed by Babcock and Bradford (2015), such an approximation is justified at recording distances that are larger than a few wavelengths.…”

Probabilistic inference of fracture-scale flow paths and aperture distribution from hydrogeophysically-monitored tracer tests

“…With FDTD it is computationally extremely demanding to model realistic fractures, for example, mm or sub-mm thin fractures in a domain of several tens of meters. Instead, most fracture-related GPR studies have relied on an analytical solution, namely the thin-bed approximation, to model EM interaction with fractures (Tsoflias and Hoch, 2006;Bradford and Deeds, 2006;Deparis and Garambois, 2008;Sassen and Everett, 2009;Sambuelli and Calzoni, 2010;Babcock and Bradford, 2015). Shakas and Linde (2015) introduced a new methodology (the so-called effective-dipole approach) to model GPR scattering from heterogeneous fractures that is inspired by a microscopic treatment of Maxwell's equations.…”

Radius estimation of subsurface cylindrical objects from ground-penetrating-radar data using full-waveform inversion

“…Gloaguen et al (2007) developed a pseudo-full-waveform inversion of borehole GPR data using stochastic tomography, whereas Cordua et al (2012) present a general Monte Carlo fullwaveform inversion strategy that integrates a priori information described by geostatistical algorithms with Bayesian inverse problem theory. Babcock and Bradford (2015) implemented the GPR waveform inversion for quantifying properties for nonaqueous phase liquid thin and ultrathin layers, and Bradford et al (2016) used a targeted GPR reflection-waveform inversion algorithm to quantify the geometry of oil spills under and within sea ice. Sassen and Everett (2009) combined the full-waveform inversion and the fully polarimetric GPR coherency technology to characterize the fractured rock, and Schmid et al (2016) studied the application of FWI to deduce the snow stratigraphy from the upward-looking GPR data.…”