guest ::
| Home > Publications database > Enhanced crosshole GPR full-waveform inversion to improve aquifer characterization |
| Home > Publications database > Enhanced crosshole GPR full-waveform inversion to improve aquifer characterization |
| Book/Dissertation / PhD Thesis | FZJ-2020-04179 |
2020
Forschungszentrum Jülich GmbH Zentralbibliothek, Verlag
Jülich
ISBN: 978-3-95806-500-0
Please use a persistent id in citations: http://hdl.handle.net/2128/26013
Abstract: Complex heterogeneities in the aquifers are critical and challenging to be detected and can have a significant effect on subsurface flow and transport. Thereby, reliable prediction of groundwater flow and solute transport is important for the protection of drinking water, and the remediation of contaminants. Small-scale high resolution images of the subsurface can help to improve the understanding of complex and heterogeneous aquifer structures that effect hydrological properties and processes. To successfully obtain hydrological parameters distributed in a 2D cross-section with high resolution, we can apply crosshole ground penetrating radar (GPR). Crosshole GPR uses high-frequency electromagnetic pulses that are emitted from a dipole-type antenna in a borehole and recorded by a receiver antenna in a second borehole. The received electromagnetic wave with its arrival time and amplitude contains information about the subsurface medium properties through which it travelled. Thereby, crosshole GPR is able to provide two electromagnetic parameters the dielectric permittivity and the electrical conductivity at the same time. Conventional inversion approaches for crosshole GPR data are generally based on geometrical ray theory, which provide relatively smooth images with a resolution that scales approximately with the diameter of the first Fresnel zone. In contrast, the crosshole GPR full-waveform inversion (FWI) provides decimeter-scalehigh resolution images, because it considers the fully recorded waveform information and the inversion is based on solving the full Maxwell’s equations. However, the crosshole GPR FWI approach also includes some limiting factors and requires several detailed processing and inversion steps. If these steps are not carefully applied, the inversion can be trapped in a local minimum. In order to minimize the influence of at least some of these factors, appreciate FWI starting models and an accurate estimation of the effective source wavelet are required. To precisely describe aquifers, high porosity layers and clay lenses, that can strongly effect flow and transport processes, need to be considered. In the framework of this thesis, we first extend this amplitude analysis approach to identify two different types of low-velocity waveguides, caused by an increased porosity and/or by a higher electrical conductivity. The obtained information about extension and dimension of such wave guiding structures is considered to improve the starting models of the FWI. Moreover, we estimate an updated effective source wavelet based on the updated permittivity starting model. To verify the presented scheme, nine GPR cross-sections were measured and analyzed at the Hermalle-sous-Argenteau test site near Liege in Belgium. Consistent structures between different cross-sections show the robustness of the updated amplitude analysis approach and the FWI results. In addition, the aquifer structures obtained from the new FWI results agree with the crosshole electrical resistivity tomography (ERT) monitoring [...]
; 
|
The record appears in these collections: |
PDF
PDF (PDFA)